bibtex_url
stringlengths 41
53
| proceedings
stringlengths 38
50
| bibtext
stringlengths 528
3.02k
| abstract
stringlengths 17
2.35k
| authors
sequencelengths 1
44
| title
stringlengths 18
190
| id
stringlengths 7
19
| arxiv_id
stringlengths 0
10
| GitHub
sequencelengths 1
1
| paper_page
stringclasses 528
values | n_linked_authors
int64 -1
15
| upvotes
int64 -1
77
| num_comments
int64 -1
10
| n_authors
int64 -1
52
| Models
sequencelengths 0
100
| Datasets
sequencelengths 0
15
| Spaces
sequencelengths 0
46
| paper_page_exists_pre_conf
int64 0
1
| type
stringclasses 2
values |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
https://aclanthology.org/2023.emnlp-main.901.bib | https://aclanthology.org/2023.emnlp-main.901/ | @inproceedings{rudman-etal-2023-outlier,
title = "Outlier Dimensions Encode Task Specific Knowledge",
author = "Rudman, William and
Chen, Catherine and
Eickhoff, Carsten",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.901",
doi = "10.18653/v1/2023.emnlp-main.901",
pages = "14596--14605",
abstract = "Representations from large language models (LLMs) are known to be dominated by a small subset of dimensions with exceedingly high variance. Previous works have argued that although ablating these outlier dimensions in LLM representations hurts downstream performance, outlier dimensions are detrimental to the representational quality of embeddings. In this study, we investigate how fine-tuning impacts outlier dimensions and show that 1) outlier dimensions that occur in pre-training persist in fine-tuned models and 2) a single outlier dimension can complete downstream tasks with a minimal error rate. Our results suggest that outlier dimensions can encode crucial task-specific knowledge and that the value of a representation in a single outlier dimension drives downstream model decisions.",
}
| Representations from large language models (LLMs) are known to be dominated by a small subset of dimensions with exceedingly high variance. Previous works have argued that although ablating these outlier dimensions in LLM representations hurts downstream performance, outlier dimensions are detrimental to the representational quality of embeddings. In this study, we investigate how fine-tuning impacts outlier dimensions and show that 1) outlier dimensions that occur in pre-training persist in fine-tuned models and 2) a single outlier dimension can complete downstream tasks with a minimal error rate. Our results suggest that outlier dimensions can encode crucial task-specific knowledge and that the value of a representation in a single outlier dimension drives downstream model decisions. | [
"Rudman, William",
"Chen, Catherine",
"Eickhoff, Carsten"
] | Outlier Dimensions Encode Task Specific Knowledge | emnlp-main.901 | [
"https://github.com/wrudman/outlier_dimensions"
] | https://huggingface.co/papers/2310.17715 | 0 | 1 | 0 | 3 | [] | [] | [] | 1 | Poster |
|
https://aclanthology.org/2023.emnlp-main.902.bib | https://aclanthology.org/2023.emnlp-main.902/ | @inproceedings{liang-etal-2023-hi,
title = "Hi-{A}r{G}: Exploring the Integration of Hierarchical Argumentation Graphs in Language Pretraining",
author = "Liang, Jingcong and
Ye, Rong and
Han, Meng and
Zhang, Qi and
Lai, Ruofei and
Zhang, Xinyu and
Cao, Zhao and
Huang, Xuanjing and
Wei, Zhongyu",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.902",
doi = "10.18653/v1/2023.emnlp-main.902",
pages = "14606--14620",
abstract = "The knowledge graph is a structure to store and represent knowledge, and recent studies have discussed its capability to assist language models for various applications. Some variations of knowledge graphs aim to record arguments and their relations for computational argumentation tasks. However, many must simplify semantic types to fit specific schemas, thus losing flexibility and expression ability. In this paper, we propose the **Hi**erarchical **Ar**gumentation **G**raph (Hi-ArG), a new structure to organize arguments. We also introduce two approaches to exploit Hi-ArG, including a text-graph multi-modal model GreaseArG and a new pre-training framework augmented with graph information. Experiments on two argumentation tasks have shown that after further pre-training and fine-tuning, GreaseArG supersedes same-scale language models on these tasks, while incorporating graph information during further pre-training can also improve the performance of vanilla language models. Code for this paper is available at {\textless}https://github.com/ljcleo/Hi-ArG{\textgreater}.",
}
| The knowledge graph is a structure to store and represent knowledge, and recent studies have discussed its capability to assist language models for various applications. Some variations of knowledge graphs aim to record arguments and their relations for computational argumentation tasks. However, many must simplify semantic types to fit specific schemas, thus losing flexibility and expression ability. In this paper, we propose the **Hi**erarchical **Ar**gumentation **G**raph (Hi-ArG), a new structure to organize arguments. We also introduce two approaches to exploit Hi-ArG, including a text-graph multi-modal model GreaseArG and a new pre-training framework augmented with graph information. Experiments on two argumentation tasks have shown that after further pre-training and fine-tuning, GreaseArG supersedes same-scale language models on these tasks, while incorporating graph information during further pre-training can also improve the performance of vanilla language models. Code for this paper is available at {\textless}https://github.com/ljcleo/Hi-ArG{\textgreater}. | [
"Liang, Jingcong",
"Ye, Rong",
"Han, Meng",
"Zhang, Qi",
"Lai, Ruofei",
"Zhang, Xinyu",
"Cao, Zhao",
"Huang, Xuanjing",
"Wei, Zhongyu"
] | Hi-ArG: Exploring the Integration of Hierarchical Argumentation Graphs in Language Pretraining | emnlp-main.902 | 2312.00874 | [
"https://github.com/ljcleo/hi-arg"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.903.bib | https://aclanthology.org/2023.emnlp-main.903/ | @inproceedings{fu-etal-2023-biomedical,
title = "Biomedical Named Entity Recognition via Dictionary-based Synonym Generalization",
author = "Fu, Zihao and
Su, Yixuan and
Meng, Zaiqiao and
Collier, Nigel",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.903",
doi = "10.18653/v1/2023.emnlp-main.903",
pages = "14621--14635",
abstract = "Biomedical named entity recognition is one of the core tasks in biomedical natural language processing (BioNLP). To tackle this task, numerous supervised/distantly supervised approaches have been proposed. Despite their remarkable success, these approaches inescapably demand laborious human effort. To alleviate the need of human effort, dictionary-based approaches have been proposed to extract named entities simply based on a given dictionary. However, one downside of existing dictionary-based approaches is that they are challenged to identify concept synonyms that are not listed in the given dictionary, which we refer as the synonym generalization problem. In this study, we propose a novel Synonym Generalization (SynGen) framework that recognizes the biomedical concepts contained in the input text using span-based predictions. In particular, SynGen introduces two regularization terms, namely, (1) a synonym distance regularizer; and (2) a noise perturbation regularizer, to minimize the synonym generalization error. To demonstrate the effectiveness of our approach, we provide a theoretical analysis of the bound of synonym generalization error. We extensively evaluate our approach on a wide range of benchmarks and the results verify that SynGen outperforms previous dictionary-based models by notable margins. Lastly, we provide a detailed analysis to further reveal the merits and inner-workings of our approach.",
}
| Biomedical named entity recognition is one of the core tasks in biomedical natural language processing (BioNLP). To tackle this task, numerous supervised/distantly supervised approaches have been proposed. Despite their remarkable success, these approaches inescapably demand laborious human effort. To alleviate the need of human effort, dictionary-based approaches have been proposed to extract named entities simply based on a given dictionary. However, one downside of existing dictionary-based approaches is that they are challenged to identify concept synonyms that are not listed in the given dictionary, which we refer as the synonym generalization problem. In this study, we propose a novel Synonym Generalization (SynGen) framework that recognizes the biomedical concepts contained in the input text using span-based predictions. In particular, SynGen introduces two regularization terms, namely, (1) a synonym distance regularizer; and (2) a noise perturbation regularizer, to minimize the synonym generalization error. To demonstrate the effectiveness of our approach, we provide a theoretical analysis of the bound of synonym generalization error. We extensively evaluate our approach on a wide range of benchmarks and the results verify that SynGen outperforms previous dictionary-based models by notable margins. Lastly, we provide a detailed analysis to further reveal the merits and inner-workings of our approach. | [
"Fu, Zihao",
"Su, Yixuan",
"Meng, Zaiqiao",
"Collier, Nigel"
] | Biomedical Named Entity Recognition via Dictionary-based Synonym Generalization | emnlp-main.903 | 2305.13066 | [
"https://github.com/fuzihaofzh/BioNER-SynGen"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.904.bib | https://aclanthology.org/2023.emnlp-main.904/ | @inproceedings{pial-skiena-2023-gnat,
title = "{GNAT}: A General Narrative Alignment Tool",
author = "Pial, Tanzir and
Skiena, Steven",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.904",
doi = "10.18653/v1/2023.emnlp-main.904",
pages = "14636--14652",
abstract = "Algorithmic sequence alignment identifies similar segments shared between pairs of documents, and is fundamental to many NLP tasks. But it is difficult to recognize similarities between distant versions of narratives such as translations and retellings, particularly for summaries and abridgements which are much shorter than the original novels. We develop a general approach to narrative alignment coupling the Smith-Waterman algorithm from bioinformatics with modern text similarity metrics. We show that the background of alignment scores fits a Gumbel distribution, enabling us to define rigorous p-values on the significance of any alignment. We apply and evaluate our general narrative alignment tool (GNAT) on four distinct problem domains differing greatly in both the relative and absolute length of documents, namely summary-to-book alignment, translated book alignment, short story alignment, and plagiarism detection{---}demonstrating the power and performance of our methods.",
}
| Algorithmic sequence alignment identifies similar segments shared between pairs of documents, and is fundamental to many NLP tasks. But it is difficult to recognize similarities between distant versions of narratives such as translations and retellings, particularly for summaries and abridgements which are much shorter than the original novels. We develop a general approach to narrative alignment coupling the Smith-Waterman algorithm from bioinformatics with modern text similarity metrics. We show that the background of alignment scores fits a Gumbel distribution, enabling us to define rigorous p-values on the significance of any alignment. We apply and evaluate our general narrative alignment tool (GNAT) on four distinct problem domains differing greatly in both the relative and absolute length of documents, namely summary-to-book alignment, translated book alignment, short story alignment, and plagiarism detection{---}demonstrating the power and performance of our methods. | [
"Pial, Tanzir",
"Skiena, Steven"
] | GNAT: A General Narrative Alignment Tool | emnlp-main.904 | 2311.03627 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.905.bib | https://aclanthology.org/2023.emnlp-main.905/ | @inproceedings{miyano-etal-2023-self,
title = "Self-Ensemble of $N$-best Generation Hypotheses by Lexically Constrained Decoding",
author = "Miyano, Ryota and
Kajiwara, Tomoyuki and
Arase, Yuki",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.905",
doi = "10.18653/v1/2023.emnlp-main.905",
pages = "14653--14661",
abstract = "We propose a method that ensembles $N$-best hypotheses to improve natural language generation. Previous studies have achieved notable improvements in generation quality by explicitly reranking $N$-best candidates. These studies assume that there exists a hypothesis of higher quality. We expand the assumption to be more practical as there exist \textit{partly} higher quality hypotheses in the $N$-best yet they may be imperfect as the entire sentences. By merging these high-quality fragments, we can obtain a higher-quality output than the single-best sentence. Specifically, we first obtain $N$-best hypotheses and conduct token-level quality estimation. We then apply tokens that should or should not be present in the final output as lexical constraints in decoding. Empirical experiments on paraphrase generation, summarisation, and constrained text generation confirm that our method outperforms the strong $N$-best reranking methods.",
}
| We propose a method that ensembles $N$-best hypotheses to improve natural language generation. Previous studies have achieved notable improvements in generation quality by explicitly reranking $N$-best candidates. These studies assume that there exists a hypothesis of higher quality. We expand the assumption to be more practical as there exist \textit{partly} higher quality hypotheses in the $N$-best yet they may be imperfect as the entire sentences. By merging these high-quality fragments, we can obtain a higher-quality output than the single-best sentence. Specifically, we first obtain $N$-best hypotheses and conduct token-level quality estimation. We then apply tokens that should or should not be present in the final output as lexical constraints in decoding. Empirical experiments on paraphrase generation, summarisation, and constrained text generation confirm that our method outperforms the strong $N$-best reranking methods. | [
"Miyano, Ryota",
"Kajiwara, Tomoyuki",
"Arase, Yuki"
] | Self-Ensemble of N-best Generation Hypotheses by Lexically Constrained Decoding | emnlp-main.905 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
||
https://aclanthology.org/2023.emnlp-main.906.bib | https://aclanthology.org/2023.emnlp-main.906/ | @inproceedings{masry-etal-2023-unichart,
title = "{U}ni{C}hart: A Universal Vision-language Pretrained Model for Chart Comprehension and Reasoning",
author = "Masry, Ahmed and
Kavehzadeh, Parsa and
Do, Xuan Long and
Hoque, Enamul and
Joty, Shafiq",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.906",
doi = "10.18653/v1/2023.emnlp-main.906",
pages = "14662--14684",
abstract = "Charts are widely used for data analysis, providing visual representations and insights into complex data. To facilitate chart-based data analysis using natural language, several downstream tasks have been introduced recently such as chart question answering and chart summarization. However, existing methods for these tasks often rely on pretraining on language or vision-language tasks, neglecting the explicit modeling of chart structures (e.g., how chart elements are related to each other). To address this, we first build a large corpus of charts covering diverse topics and visual styles. We then present UniChart, a pretrained model for chart comprehension and reasoning. UniChart encodes the relevant text, data, and visual elements of charts and then uses a chart-grounded text decoder for text generation. We propose several chart-specific pretraining tasks that include: (i) low-level tasks to extract the visual elements (e.g., bars, lines) and data from charts, and (ii) high-level tasks to acquire chart understanding and reasoning skills. Our experiments demonstrate that pretraining UniChart on a large corpus with chart-specific objectives, followed by fine-tuning, yields state-of-the-art performance on four downstream tasks. Moreover, our model exhibits superior generalizability to unseen chart corpus, surpassing previous approaches that lack chart-specific objectives and utilize limited chart resources.",
}
| Charts are widely used for data analysis, providing visual representations and insights into complex data. To facilitate chart-based data analysis using natural language, several downstream tasks have been introduced recently such as chart question answering and chart summarization. However, existing methods for these tasks often rely on pretraining on language or vision-language tasks, neglecting the explicit modeling of chart structures (e.g., how chart elements are related to each other). To address this, we first build a large corpus of charts covering diverse topics and visual styles. We then present UniChart, a pretrained model for chart comprehension and reasoning. UniChart encodes the relevant text, data, and visual elements of charts and then uses a chart-grounded text decoder for text generation. We propose several chart-specific pretraining tasks that include: (i) low-level tasks to extract the visual elements (e.g., bars, lines) and data from charts, and (ii) high-level tasks to acquire chart understanding and reasoning skills. Our experiments demonstrate that pretraining UniChart on a large corpus with chart-specific objectives, followed by fine-tuning, yields state-of-the-art performance on four downstream tasks. Moreover, our model exhibits superior generalizability to unseen chart corpus, surpassing previous approaches that lack chart-specific objectives and utilize limited chart resources. | [
"Masry, Ahmed",
"Kavehzadeh, Parsa",
"Do, Xuan Long",
"Hoque, Enamul",
"Joty, Shafiq"
] | UniChart: A Universal Vision-language Pretrained Model for Chart Comprehension and Reasoning | emnlp-main.906 | 2305.14761 | [
"https://github.com/vis-nlp/unichart"
] | https://huggingface.co/papers/2305.14761 | 1 | 0 | 0 | 5 | [
"ahmed-masry/unichart-base-960"
] | [
"ahmed-masry/unichart-pretrain-data"
] | [
"ahmed-masry/UniChart"
] | 1 | Poster |
https://aclanthology.org/2023.emnlp-main.907.bib | https://aclanthology.org/2023.emnlp-main.907/ | @inproceedings{he-etal-2023-merging,
title = "Merging Experts into One: Improving Computational Efficiency of Mixture of Experts",
author = "He, Shwai and
Fan, Run-Ze and
Ding, Liang and
Shen, Li and
Zhou, Tianyi and
Tao, Dacheng",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.907",
doi = "10.18653/v1/2023.emnlp-main.907",
pages = "14685--14691",
abstract = "Scaling the size of language models usually leads to remarkable advancements in NLP tasks. But it often comes with a price of growing computational cost. Although a sparse Mixture of Experts (MoE) can reduce the cost by activating a small subset of parameters (e.g., one expert) for each input, its computation escalates significantly if increasing the number of activated experts, limiting its practical utility. Can we retain the advantages of adding more experts without substantially increasing the computational costs? In this paper, we first demonstrate the superiority of selecting multiple experts and then propose a computation-efficient approach called \textbf{Merging Experts into One} (MEO), which reduces the computation cost to that of a single expert. Extensive experiments show that MEO significantly improves computational efficiency, e.g., FLOPS drops from 72.0G of vanilla MoE to 28.6G (MEO). Moreover, we propose a token-level attention block that further enhances the efficiency and performance of token-level MEO, e.g., 83.3{\%} (MEO) vs. 82.6{\%} (vanilla MoE) average score on the GLUE benchmark. Our code will be released upon acceptance. Code will be released at: \url{https://github.com/Shwai-He/MEO}.",
}
| Scaling the size of language models usually leads to remarkable advancements in NLP tasks. But it often comes with a price of growing computational cost. Although a sparse Mixture of Experts (MoE) can reduce the cost by activating a small subset of parameters (e.g., one expert) for each input, its computation escalates significantly if increasing the number of activated experts, limiting its practical utility. Can we retain the advantages of adding more experts without substantially increasing the computational costs? In this paper, we first demonstrate the superiority of selecting multiple experts and then propose a computation-efficient approach called \textbf{Merging Experts into One} (MEO), which reduces the computation cost to that of a single expert. Extensive experiments show that MEO significantly improves computational efficiency, e.g., FLOPS drops from 72.0G of vanilla MoE to 28.6G (MEO). Moreover, we propose a token-level attention block that further enhances the efficiency and performance of token-level MEO, e.g., 83.3{\%} (MEO) vs. 82.6{\%} (vanilla MoE) average score on the GLUE benchmark. Our code will be released upon acceptance. Code will be released at: \url{https://github.com/Shwai-He/MEO}. | [
"He, Shwai",
"Fan, Run-Ze",
"Ding, Liang",
"Shen, Li",
"Zhou, Tianyi",
"Tao, Dacheng"
] | Merging Experts into One: Improving Computational Efficiency of Mixture of Experts | emnlp-main.907 | 2310.09832 | [
"https://github.com/shwai-he/meo"
] | https://huggingface.co/papers/2310.09832 | 4 | 1 | 0 | 6 | [] | [] | [] | 1 | Oral |
https://aclanthology.org/2023.emnlp-main.908.bib | https://aclanthology.org/2023.emnlp-main.908/ | @inproceedings{shomer-etal-2023-distance,
title = "Distance-Based Propagation for Efficient Knowledge Graph Reasoning",
author = "Shomer, Harry and
Ma, Yao and
Li, Juanhui and
Wu, Bo and
Aggarwal, Charu and
Tang, Jiliang",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.908",
doi = "10.18653/v1/2023.emnlp-main.908",
pages = "14692--14707",
abstract = "Knowledge graph completion (KGC) aims to predict unseen edges in knowledge graphs (KGs), resulting in the discovery of new facts. A new class of methods have been proposed to tackle this problem by aggregating path information. These methods have shown tremendous ability in the task of KGC. However they are plagued by efficiency issues. Though there are a few recent attempts to address this through learnable path pruning, they often sacrifice the performance to gain efficiency. In this work, we identify two intrinsic limitations of these methods that affect the efficiency and representation quality. To address the limitations, we introduce a new method, TAGNet, which is able to efficiently propagate information. This is achieved by only aggregating paths in a fixed window for each source-target pair. We demonstrate that the complexity of TAGNet is independent of the number of layers. Extensive experiments demonstrate that TAGNet can cut down on the number of propagated messages by as much as 90{\%} while achieving competitive performance on multiple KG datasets.",
}
| Knowledge graph completion (KGC) aims to predict unseen edges in knowledge graphs (KGs), resulting in the discovery of new facts. A new class of methods have been proposed to tackle this problem by aggregating path information. These methods have shown tremendous ability in the task of KGC. However they are plagued by efficiency issues. Though there are a few recent attempts to address this through learnable path pruning, they often sacrifice the performance to gain efficiency. In this work, we identify two intrinsic limitations of these methods that affect the efficiency and representation quality. To address the limitations, we introduce a new method, TAGNet, which is able to efficiently propagate information. This is achieved by only aggregating paths in a fixed window for each source-target pair. We demonstrate that the complexity of TAGNet is independent of the number of layers. Extensive experiments demonstrate that TAGNet can cut down on the number of propagated messages by as much as 90{\%} while achieving competitive performance on multiple KG datasets. | [
"Shomer, Harry",
"Ma, Yao",
"Li, Juanhui",
"Wu, Bo",
"Aggarwal, Charu",
"Tang, Jiliang"
] | Distance-Based Propagation for Efficient Knowledge Graph Reasoning | emnlp-main.908 | 2311.01024 | [
"https://github.com/harryshomer/tagnet"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.909.bib | https://aclanthology.org/2023.emnlp-main.909/ | @inproceedings{sancheti-etal-2023-read,
title = "What to Read in a Contract? Party-Specific Summarization of Legal Obligations, Entitlements, and Prohibitions",
author = "Sancheti, Abhilasha and
Garimella, Aparna and
Srinivasan, Balaji and
Rudinger, Rachel",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.909",
doi = "10.18653/v1/2023.emnlp-main.909",
pages = "14708--14725",
abstract = "Reviewing and comprehending key obligations, entitlements, and prohibitions in legal contracts can be a tedious task due to their length and domain-specificity. Furthermore, the key rights and duties requiring review vary for each contracting party. In this work, we propose a new task of \textit{party-specific} extractive summarization for legal contracts to facilitate faster reviewing and improved comprehension of rights and duties. To facilitate this, we curate a dataset comprising of party-specific pairwise importance comparisons annotated by legal experts, covering {\textasciitilde}293K sentence pairs that include obligations, entitlements, and prohibitions extracted from lease agreements. Using this dataset, we train a pairwise importance ranker and propose a pipeline-based extractive summarization system that generates a party-specific contract summary. We establish the need for incorporating domain-specific notions of importance during summarization by comparing our system against various baselines using both automatic and human evaluation methods.",
}
| Reviewing and comprehending key obligations, entitlements, and prohibitions in legal contracts can be a tedious task due to their length and domain-specificity. Furthermore, the key rights and duties requiring review vary for each contracting party. In this work, we propose a new task of \textit{party-specific} extractive summarization for legal contracts to facilitate faster reviewing and improved comprehension of rights and duties. To facilitate this, we curate a dataset comprising of party-specific pairwise importance comparisons annotated by legal experts, covering {\textasciitilde}293K sentence pairs that include obligations, entitlements, and prohibitions extracted from lease agreements. Using this dataset, we train a pairwise importance ranker and propose a pipeline-based extractive summarization system that generates a party-specific contract summary. We establish the need for incorporating domain-specific notions of importance during summarization by comparing our system against various baselines using both automatic and human evaluation methods. | [
"Sancheti, Abhilasha",
"Garimella, Aparna",
"Srinivasan, Balaji",
"Rudinger, Rachel"
] | What to Read in a Contract? Party-Specific Summarization of Legal Obligations, Entitlements, and Prohibitions | emnlp-main.909 | 2212.09825 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.910.bib | https://aclanthology.org/2023.emnlp-main.910/ | @inproceedings{lee-etal-2023-enhancing-computation,
title = "Enhancing Computation Efficiency in Large Language Models through Weight and Activation Quantization",
author = "Lee, Janghwan and
Kim, Minsoo and
Baek, Seungcheol and
Hwang, Seok and
Sung, Wonyong and
Choi, Jungwook",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.910",
doi = "10.18653/v1/2023.emnlp-main.910",
pages = "14726--14739",
abstract = "Large Language Models (LLMs) are proficient in natural language processing tasks, but their deployment is often restricted by extensive parameter sizes and computational demands. This paper focuses on post-training quantization (PTQ) in LLMs, specifically 4-bit weight and 8-bit activation (W4A8) quantization, to enhance computational efficiency{---}a topic less explored compared to weight-only quantization. We present two innovative techniques: activation-quantization-aware scaling (AQAS) and sequence-length-aware calibration (SLAC) to enhance PTQ by considering the combined effects on weights and activations and aligning calibration sequence lengths to target tasks. Moreover, we introduce dINT, a hybrid data format combining integer and denormal representations, to address the underflow issue in W4A8 quantization, where small values are rounded to zero. Through rigorous evaluations of LLMs, including OPT and LLaMA, we demonstrate that our techniques significantly boost task accuracies to levels comparable with full-precision models. By developing arithmetic units compatible with dINT, we further confirm that our methods yield a 2$\times$ hardware efficiency improvement compared to 8-bit integer MAC unit.",
}
| Large Language Models (LLMs) are proficient in natural language processing tasks, but their deployment is often restricted by extensive parameter sizes and computational demands. This paper focuses on post-training quantization (PTQ) in LLMs, specifically 4-bit weight and 8-bit activation (W4A8) quantization, to enhance computational efficiency{---}a topic less explored compared to weight-only quantization. We present two innovative techniques: activation-quantization-aware scaling (AQAS) and sequence-length-aware calibration (SLAC) to enhance PTQ by considering the combined effects on weights and activations and aligning calibration sequence lengths to target tasks. Moreover, we introduce dINT, a hybrid data format combining integer and denormal representations, to address the underflow issue in W4A8 quantization, where small values are rounded to zero. Through rigorous evaluations of LLMs, including OPT and LLaMA, we demonstrate that our techniques significantly boost task accuracies to levels comparable with full-precision models. By developing arithmetic units compatible with dINT, we further confirm that our methods yield a 2$\times$ hardware efficiency improvement compared to 8-bit integer MAC unit. | [
"Lee, Janghwan",
"Kim, Minsoo",
"Baek, Seungcheol",
"Hwang, Seok",
"Sung, Wonyong",
"Choi, Jungwook"
] | Enhancing Computation Efficiency in Large Language Models through Weight and Activation Quantization | emnlp-main.910 | 2311.05161 | [
""
] | https://huggingface.co/papers/2311.05161 | 0 | 0 | 0 | 6 | [] | [] | [] | 1 | Poster |
https://aclanthology.org/2023.emnlp-main.911.bib | https://aclanthology.org/2023.emnlp-main.911/ | @inproceedings{ye-etal-2023-cp,
title = "{CP}-{BCS}: Binary Code Summarization Guided by Control Flow Graph and Pseudo Code",
author = "Ye, Tong and
Wu, Lingfei and
Ma, Tengfei and
Zhang, Xuhong and
Du, Yangkai and
Liu, Peiyu and
Ji, Shouling and
Wang, Wenhai",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.911",
doi = "10.18653/v1/2023.emnlp-main.911",
pages = "14740--14752",
abstract = "Automatically generating function summaries for binaries is an extremely valuable but challenging task, since it involves translating the execution behavior and semantics of the low-level language (assembly code) into human-readable natural language. However, most current works on understanding assembly code are oriented towards generating function names, which involve numerous abbreviations that make them still confusing. To bridge this gap, we focus on generating complete summaries for binary functions, especially for stripped binary (no symbol table and debug information in reality). To fully exploit the semantics of assembly code, we present a control flow graph and pseudo code guided binary code summarization framework called CP-BCS. CP-BCS utilizes a bidirectional instruction-level control flow graph and pseudo code that incorporates expert knowledge to learn the comprehensive binary function execution behavior and logic semantics. We evaluate CP-BCS on 3 different binary optimization levels (O1, O2, and O3) for 3 different computer architectures (X86, X64, and ARM). The evaluation results demonstrate CP-BCS is superior and significantly improves the efficiency of reverse engineering.",
}
| Automatically generating function summaries for binaries is an extremely valuable but challenging task, since it involves translating the execution behavior and semantics of the low-level language (assembly code) into human-readable natural language. However, most current works on understanding assembly code are oriented towards generating function names, which involve numerous abbreviations that make them still confusing. To bridge this gap, we focus on generating complete summaries for binary functions, especially for stripped binary (no symbol table and debug information in reality). To fully exploit the semantics of assembly code, we present a control flow graph and pseudo code guided binary code summarization framework called CP-BCS. CP-BCS utilizes a bidirectional instruction-level control flow graph and pseudo code that incorporates expert knowledge to learn the comprehensive binary function execution behavior and logic semantics. We evaluate CP-BCS on 3 different binary optimization levels (O1, O2, and O3) for 3 different computer architectures (X86, X64, and ARM). The evaluation results demonstrate CP-BCS is superior and significantly improves the efficiency of reverse engineering. | [
"Ye, Tong",
"Wu, Lingfei",
"Ma, Tengfei",
"Zhang, Xuhong",
"Du, Yangkai",
"Liu, Peiyu",
"Ji, Shouling",
"Wang, Wenhai"
] | CP-BCS: Binary Code Summarization Guided by Control Flow Graph and Pseudo Code | emnlp-main.911 | 2310.16853 | [
"https://github.com/tongye98/binarycodesummary"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.912.bib | https://aclanthology.org/2023.emnlp-main.912/ | @inproceedings{ye-etal-2023-assessing,
title = "Assessing Step-by-Step Reasoning against Lexical Negation: A Case Study on Syllogism",
author = "Ye, Mengyu and
Kuribayashi, Tatsuki and
Suzuki, Jun and
Kobayashi, Goro and
Funayama, Hiroaki",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.912",
doi = "10.18653/v1/2023.emnlp-main.912",
pages = "14753--14773",
abstract = "Large language models (LLMs) take advantage of step-by-step reasoning instructions, e.g., chain-of-thought (CoT) prompting. Building on this, their ability to perform CoT-style reasoning robustly is of interest from a probing perspective. In this study, we inspect the step-by-step reasoning ability of LLMs with a focus on negation, which is a core linguistic phenomenon that is difficult to process. In particular, we introduce several controlled settings (e.g., reasoning in case of fictional entities) to evaluate the logical reasoning abilities of the models. We observed that dozens of modern LLMs were not robust against lexical negation (e.g., plausible$\rightarrow$implausible) when performing CoT-style reasoning, and the results highlight unique limitations in each LLM family.",
}
| Large language models (LLMs) take advantage of step-by-step reasoning instructions, e.g., chain-of-thought (CoT) prompting. Building on this, their ability to perform CoT-style reasoning robustly is of interest from a probing perspective. In this study, we inspect the step-by-step reasoning ability of LLMs with a focus on negation, which is a core linguistic phenomenon that is difficult to process. In particular, we introduce several controlled settings (e.g., reasoning in case of fictional entities) to evaluate the logical reasoning abilities of the models. We observed that dozens of modern LLMs were not robust against lexical negation (e.g., plausible$\rightarrow$implausible) when performing CoT-style reasoning, and the results highlight unique limitations in each LLM family. | [
"Ye, Mengyu",
"Kuribayashi, Tatsuki",
"Suzuki, Jun",
"Kobayashi, Goro",
"Funayama, Hiroaki"
] | Assessing Step-by-Step Reasoning against Lexical Negation: A Case Study on Syllogism | emnlp-main.912 | 2310.14868 | [
"https://github.com/muyo8692/stepbystep-reasoning-vs-negation"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Oral |
|
https://aclanthology.org/2023.emnlp-main.913.bib | https://aclanthology.org/2023.emnlp-main.913/ | @inproceedings{fan-etal-2023-chain,
title = "Chain-of-Thought Tuning: Masked Language Models can also Think Step By Step in Natural Language Understanding",
author = "Fan, Caoyun and
Tian, Jidong and
Li, Yitian and
Chen, Wenqing and
He, Hao and
Jin, Yaohui",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.913",
doi = "10.18653/v1/2023.emnlp-main.913",
pages = "14774--14785",
abstract = "Chain-of-Thought (CoT) is a technique that guides Large Language Models (LLMs) to decompose complex tasks into multi-step reasoning through intermediate steps in natural language form. Briefly, CoT enables LLMs to think step by step. However, although many Natural Language Understanding (NLU) tasks also require thinking step by step, LLMs perform less well than small-scale Masked Language Models (MLMs). To migrate CoT from LLMs to MLMs, we propose Chain-of-Thought Tuning (CoTT), a two-step reasoning framework based on prompt tuning, to implement step-by-step thinking for MLMs on NLU tasks. From the perspective of CoT, CoTT{'}s two-step framework enables MLMs to implement task decomposition; CoTT{'}s prompt tuning allows intermediate steps to be used in natural language form. Thereby, the success of CoT can be extended to NLU tasks through MLMs. To verify the effectiveness of CoTT, we conduct experiments on two NLU tasks: hierarchical classification and relation extraction, and the results show that CoTT outperforms baselines and achieves state-of-the-art performance.",
}
| Chain-of-Thought (CoT) is a technique that guides Large Language Models (LLMs) to decompose complex tasks into multi-step reasoning through intermediate steps in natural language form. Briefly, CoT enables LLMs to think step by step. However, although many Natural Language Understanding (NLU) tasks also require thinking step by step, LLMs perform less well than small-scale Masked Language Models (MLMs). To migrate CoT from LLMs to MLMs, we propose Chain-of-Thought Tuning (CoTT), a two-step reasoning framework based on prompt tuning, to implement step-by-step thinking for MLMs on NLU tasks. From the perspective of CoT, CoTT{'}s two-step framework enables MLMs to implement task decomposition; CoTT{'}s prompt tuning allows intermediate steps to be used in natural language form. Thereby, the success of CoT can be extended to NLU tasks through MLMs. To verify the effectiveness of CoTT, we conduct experiments on two NLU tasks: hierarchical classification and relation extraction, and the results show that CoTT outperforms baselines and achieves state-of-the-art performance. | [
"Fan, Caoyun",
"Tian, Jidong",
"Li, Yitian",
"Chen, Wenqing",
"He, Hao",
"Jin, Yaohui"
] | Chain-of-Thought Tuning: Masked Language Models can also Think Step By Step in Natural Language Understanding | emnlp-main.913 | 2310.11721 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.914.bib | https://aclanthology.org/2023.emnlp-main.914/ | @inproceedings{zhao-etal-2023-large,
title = "Large Language Models are Complex Table Parsers",
author = "Zhao, Bowen and
Ji, Changkai and
Zhang, Yuejie and
He, Wen and
Wang, Yingwen and
Wang, Qing and
Feng, Rui and
Zhang, Xiaobo",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.914",
doi = "10.18653/v1/2023.emnlp-main.914",
pages = "14786--14802",
abstract = "With the Generative Pre-trained Transformer 3.5 (GPT-3.5) exhibiting remarkable reasoning and comprehension abilities in Natural Language Processing (NLP), most Question Answering (QA) research has primarily centered around general QA tasks based on GPT, neglecting the specific challenges posed by Complex Table QA. In this paper, we propose to incorporate GPT-3.5 to address such challenges, in which complex tables are reconstructed into tuples and specific prompt designs are employed for dialogues. Specifically, we encode each cell{'}s hierarchical structure, position information, and content as a tuple. By enhancing the prompt template with an explanatory description of the meaning of each tuple and the logical reasoning process of the task, we effectively improve the hierarchical structure awareness capability of GPT-3.5 to better parse the complex tables. Extensive experiments and results on Complex Table QA datasets, i.e., the open-domain dataset HiTAB and the aviation domain dataset AIT-QA show that our approach significantly outperforms previous work on both datasets, leading to state-of-the-art (SOTA) performance.",
}
| With the Generative Pre-trained Transformer 3.5 (GPT-3.5) exhibiting remarkable reasoning and comprehension abilities in Natural Language Processing (NLP), most Question Answering (QA) research has primarily centered around general QA tasks based on GPT, neglecting the specific challenges posed by Complex Table QA. In this paper, we propose to incorporate GPT-3.5 to address such challenges, in which complex tables are reconstructed into tuples and specific prompt designs are employed for dialogues. Specifically, we encode each cell{'}s hierarchical structure, position information, and content as a tuple. By enhancing the prompt template with an explanatory description of the meaning of each tuple and the logical reasoning process of the task, we effectively improve the hierarchical structure awareness capability of GPT-3.5 to better parse the complex tables. Extensive experiments and results on Complex Table QA datasets, i.e., the open-domain dataset HiTAB and the aviation domain dataset AIT-QA show that our approach significantly outperforms previous work on both datasets, leading to state-of-the-art (SOTA) performance. | [
"Zhao, Bowen",
"Ji, Changkai",
"Zhang, Yuejie",
"He, Wen",
"Wang, Yingwen",
"Wang, Qing",
"Feng, Rui",
"Zhang, Xiaobo"
] | Large Language Models are Complex Table Parsers | emnlp-main.914 | 2312.11521 | [
""
] | https://huggingface.co/papers/2312.11521 | 0 | 0 | 0 | 8 | [] | [] | [] | 1 | Poster |
https://aclanthology.org/2023.emnlp-main.915.bib | https://aclanthology.org/2023.emnlp-main.915/ | @inproceedings{fan-etal-2023-r2h,
title = "{R}2{H}: Building Multimodal Navigation Helpers that Respond to Help Requests",
author = "Fan, Yue and
Gu, Jing and
Zheng, Kaizhi and
Wang, Xin",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.915",
doi = "10.18653/v1/2023.emnlp-main.915",
pages = "14803--14819",
abstract = "Intelligent navigation-helper agents are critical as they can navigate users in unknown areas through environmental awareness and conversational ability, serving as potential accessibility tools for individuals with disabilities. In this work, we first introduce a novel benchmark, Respond to Help Requests (R2H), to promote the development of multi-modal navigation helpers capable of responding to requests for help, utilizing existing dialog-based embodied datasets. R2H mainly includes two tasks: (1) Respond to Dialog History (RDH), which assesses the helper agent{'}s ability to generate informative responses based on a given dialog history, and (2) Respond during Interaction (RdI), which evaluates the effectiveness and efficiency of the response during consistent cooperation with a task performer. Furthermore, we explore two approaches to construct the navigation-helper agent, including fine-tuning a novel task-oriented multi-modal response generation model that can see and respond, named SeeRee, and employing a multi-modal large language model in a zero-shot manner. Analysis of the task and method was conducted based on both automatic benchmarking and human evaluations.",
}
| Intelligent navigation-helper agents are critical as they can navigate users in unknown areas through environmental awareness and conversational ability, serving as potential accessibility tools for individuals with disabilities. In this work, we first introduce a novel benchmark, Respond to Help Requests (R2H), to promote the development of multi-modal navigation helpers capable of responding to requests for help, utilizing existing dialog-based embodied datasets. R2H mainly includes two tasks: (1) Respond to Dialog History (RDH), which assesses the helper agent{'}s ability to generate informative responses based on a given dialog history, and (2) Respond during Interaction (RdI), which evaluates the effectiveness and efficiency of the response during consistent cooperation with a task performer. Furthermore, we explore two approaches to construct the navigation-helper agent, including fine-tuning a novel task-oriented multi-modal response generation model that can see and respond, named SeeRee, and employing a multi-modal large language model in a zero-shot manner. Analysis of the task and method was conducted based on both automatic benchmarking and human evaluations. | [
"Fan, Yue",
"Gu, Jing",
"Zheng, Kaizhi",
"Wang, Xin"
] | R2H: Building Multimodal Navigation Helpers that Respond to Help Requests | emnlp-main.915 | 2305.14260 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.916.bib | https://aclanthology.org/2023.emnlp-main.916/ | @inproceedings{mittal-etal-2023-speech,
title = "Speech-enriched Memory for Inference-time Adaptation of {ASR} Models to Word Dictionaries",
author = "Mittal, Ashish and
Sarawagi, Sunita and
Jyothi, Preethi and
Saon, George and
Kurata, Gakuto",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.916",
doi = "10.18653/v1/2023.emnlp-main.916",
pages = "14820--14835",
abstract = "Despite the impressive performance of ASR models on mainstream benchmarks, their performance on rare words is unsatisfactory. In enterprise settings, often a focused list of entities (such as locations, names, etc) are available which can be used to adapt the model to the terminology of specific domains. In this paper, we present a novel inference algorithm that improves the prediction of state-of-the-art ASR models using nearest-neighbor-based matching on an inference-time word list. We consider both the Transducer architecture that is useful in the streaming setting, and state-of-the-art encoder-decoder models such as Whisper. In our approach, a list of rare entities is indexed in a memory by synthesizing speech for each entry, and then storing the internal acoustic and language model states obtained from the best possible alignment on the ASR model. The memory is organized as a trie which we harness to perform a stateful lookup during inference. A key property of our extension is that we prevent spurious matches by restricting to only word-level matches. In our experiments on publicly available datasets and private benchmarks, we show that our method is effective in significantly improving rare word recognition.",
}
| Despite the impressive performance of ASR models on mainstream benchmarks, their performance on rare words is unsatisfactory. In enterprise settings, often a focused list of entities (such as locations, names, etc) are available which can be used to adapt the model to the terminology of specific domains. In this paper, we present a novel inference algorithm that improves the prediction of state-of-the-art ASR models using nearest-neighbor-based matching on an inference-time word list. We consider both the Transducer architecture that is useful in the streaming setting, and state-of-the-art encoder-decoder models such as Whisper. In our approach, a list of rare entities is indexed in a memory by synthesizing speech for each entry, and then storing the internal acoustic and language model states obtained from the best possible alignment on the ASR model. The memory is organized as a trie which we harness to perform a stateful lookup during inference. A key property of our extension is that we prevent spurious matches by restricting to only word-level matches. In our experiments on publicly available datasets and private benchmarks, we show that our method is effective in significantly improving rare word recognition. | [
"Mittal, Ashish",
"Sarawagi, Sunita",
"Jyothi, Preethi",
"Saon, George",
"Kurata, Gakuto"
] | Speech-enriched Memory for Inference-time Adaptation of ASR Models to Word Dictionaries | emnlp-main.916 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
||
https://aclanthology.org/2023.emnlp-main.917.bib | https://aclanthology.org/2023.emnlp-main.917/ | @inproceedings{zhang-etal-2023-generative,
title = "Generative Table Pre-training Empowers Models for Tabular Prediction",
author = "Zhang, Tianping and
Wang, Shaowen and
Yan, Shuicheng and
Jian, Li and
Liu, Qian",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.917",
doi = "10.18653/v1/2023.emnlp-main.917",
pages = "14836--14854",
abstract = "Recently, the topic of table pre-training has attracted considerable research interest. However, how to employ table pre-training to boost the performance of tabular prediction remains an open challenge. In this paper, we propose TapTap, the first attempt that leverages table pre-training to empower models for tabular prediction. After pre-training on a large corpus of real-world tabular data, TapTap can generate high-quality synthetic tables to support various applications on tabular data, including privacy protection, low resource regime, missing value imputation, and imbalanced classification. Extensive experiments on 12 datasets demonstrate that TapTap outperforms a total of 16 baselines in different scenarios. Meanwhile, it can be easily combined with various backbone models, including LightGBM, Multilayer Perceptron (MLP) and Transformer. Moreover, with the aid of table pre-training, models trained using synthetic data generated by TapTap can even compete with models using the original dataset on half of the experimental datasets, marking a milestone in the development of synthetic tabular data generation. The code and datasets are available at https://github.com/ZhangTP1996/TapTap.",
}
| Recently, the topic of table pre-training has attracted considerable research interest. However, how to employ table pre-training to boost the performance of tabular prediction remains an open challenge. In this paper, we propose TapTap, the first attempt that leverages table pre-training to empower models for tabular prediction. After pre-training on a large corpus of real-world tabular data, TapTap can generate high-quality synthetic tables to support various applications on tabular data, including privacy protection, low resource regime, missing value imputation, and imbalanced classification. Extensive experiments on 12 datasets demonstrate that TapTap outperforms a total of 16 baselines in different scenarios. Meanwhile, it can be easily combined with various backbone models, including LightGBM, Multilayer Perceptron (MLP) and Transformer. Moreover, with the aid of table pre-training, models trained using synthetic data generated by TapTap can even compete with models using the original dataset on half of the experimental datasets, marking a milestone in the development of synthetic tabular data generation. The code and datasets are available at https://github.com/ZhangTP1996/TapTap. | [
"Zhang, Tianping",
"Wang, Shaowen",
"Yan, Shuicheng",
"Jian, Li",
"Liu, Qian"
] | Generative Table Pre-training Empowers Models for Tabular Prediction | emnlp-main.917 | 2305.09696 | [
"https://github.com/zhangtp1996/taptap"
] | https://huggingface.co/papers/2305.09696 | 1 | 0 | 0 | 5 | [] | [
"ztphs980/taptap_datasets"
] | [] | 1 | Poster |
https://aclanthology.org/2023.emnlp-main.918.bib | https://aclanthology.org/2023.emnlp-main.918/ | @inproceedings{zhu-etal-2023-learning,
title = "Learning to Describe for Predicting Zero-shot Drug-Drug Interactions",
author = "Zhu, Fangqi and
Zhang, Yongqi and
Chen, Lei and
Qin, Bing and
Xu, Ruifeng",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.918",
doi = "10.18653/v1/2023.emnlp-main.918",
pages = "14855--14870",
abstract = "Adverse drug-drug interactions (DDIs) can compromise the effectiveness of concurrent drug administration, posing a significant challenge in healthcare. As the development of new drugs continues, the potential for unknown adverse effects resulting from DDIs becomes a growing concern. Traditional computational methods for DDI prediction may fail to capture interactions for new drugs due to the lack of knowledge. In this paper, we introduce a new problem setup as zero-shot DDI prediction that deals with the case of new drugs. Leveraging textual information from online databases like DrugBank and PubChem, we propose an innovative approach TextDDI with a language model-based DDI predictor and a reinforcement learning (RL)-based information selector, enabling the selection of concise and pertinent text for accurate DDI prediction on new drugs. Empirical results show the benefits of the proposed approach on several settings including zero-shot and few-shot DDI prediction, and the selected texts are semantically relevant. Our code and data are available at https://github.com/zhufq00/DDIs-Prediction.",
}
| Adverse drug-drug interactions (DDIs) can compromise the effectiveness of concurrent drug administration, posing a significant challenge in healthcare. As the development of new drugs continues, the potential for unknown adverse effects resulting from DDIs becomes a growing concern. Traditional computational methods for DDI prediction may fail to capture interactions for new drugs due to the lack of knowledge. In this paper, we introduce a new problem setup as zero-shot DDI prediction that deals with the case of new drugs. Leveraging textual information from online databases like DrugBank and PubChem, we propose an innovative approach TextDDI with a language model-based DDI predictor and a reinforcement learning (RL)-based information selector, enabling the selection of concise and pertinent text for accurate DDI prediction on new drugs. Empirical results show the benefits of the proposed approach on several settings including zero-shot and few-shot DDI prediction, and the selected texts are semantically relevant. Our code and data are available at https://github.com/zhufq00/DDIs-Prediction. | [
"Zhu, Fangqi",
"Zhang, Yongqi",
"Chen, Lei",
"Qin, Bing",
"Xu, Ruifeng"
] | Learning to Describe for Predicting Zero-shot Drug-Drug Interactions | emnlp-main.918 | 2403.08377 | [
"https://github.com/zhufq00/ddis-prediction"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Oral |
|
https://aclanthology.org/2023.emnlp-main.919.bib | https://aclanthology.org/2023.emnlp-main.919/ | @inproceedings{xenos-etal-2023-simple,
title = "A Simple Baseline for Knowledge-Based Visual Question Answering",
author = "Xenos, Alexandros and
Stafylakis, Themos and
Patras, Ioannis and
Tzimiropoulos, Georgios",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.919",
doi = "10.18653/v1/2023.emnlp-main.919",
pages = "14871--14877",
abstract = "This paper is on the problem of Knowledge-Based Visual Question Answering (KB-VQA). Recent works have emphasized the significance of incorporating both explicit (through external databases) and implicit (through LLMs) knowledge to answer questions requiring external knowledge effectively. A common limitation of such approaches is that they consist of relatively complicated pipelines and often heavily rely on accessing GPT-3 API. Our main contribution in this paper is to propose a much simpler and readily reproducible pipeline which, in a nutshell, is based on efficient in-context learning by prompting LLaMA (1 and 2) using question-informative captions as contextual information. Contrary to recent approaches, our method is training-free, does not require access to external databases or APIs, and yet achieves state-of-the-art accuracy on the OK-VQA and A-OK-VQA datasets. Finally, we perform several ablation studies to understand important aspects of our method. Our code is publicly available at https://github.com/alexandrosXe/ASimple-Baseline-For-Knowledge-Based-VQA",
}
| This paper is on the problem of Knowledge-Based Visual Question Answering (KB-VQA). Recent works have emphasized the significance of incorporating both explicit (through external databases) and implicit (through LLMs) knowledge to answer questions requiring external knowledge effectively. A common limitation of such approaches is that they consist of relatively complicated pipelines and often heavily rely on accessing GPT-3 API. Our main contribution in this paper is to propose a much simpler and readily reproducible pipeline which, in a nutshell, is based on efficient in-context learning by prompting LLaMA (1 and 2) using question-informative captions as contextual information. Contrary to recent approaches, our method is training-free, does not require access to external databases or APIs, and yet achieves state-of-the-art accuracy on the OK-VQA and A-OK-VQA datasets. Finally, we perform several ablation studies to understand important aspects of our method. Our code is publicly available at https://github.com/alexandrosXe/ASimple-Baseline-For-Knowledge-Based-VQA | [
"Xenos, Alex",
"ros",
"Stafylakis, Themos",
"Patras, Ioannis",
"Tzimiropoulos, Georgios"
] | A Simple Baseline for Knowledge-Based Visual Question Answering | emnlp-main.919 | 2310.13570 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.920.bib | https://aclanthology.org/2023.emnlp-main.920/ | @inproceedings{mahbub-etal-2023-unveiling,
title = "Unveiling the Essence of Poetry: Introducing a Comprehensive Dataset and Benchmark for Poem Summarization",
author = "Mahbub, Ridwan and
Khan, Ifrad and
Anuva, Samiha and
Shahriar, Md Shihab and
Laskar, Md Tahmid Rahman and
Ahmed, Sabbir",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.920",
doi = "10.18653/v1/2023.emnlp-main.920",
pages = "14878--14886",
abstract = "While research in natural language processing has progressed significantly in creative language generation, the question of whether language models can interpret the intended meaning of creative language largely remains unanswered. Poetry as a creative art form has existed for generations, and summarization of such content requires deciphering the figurative patterns to find out the actual intent and message of the poet. This task can provide the researchers an opportunity to evaluate the creative language interpretation capacity of the language models. Unlike typical text, summarization of poems is a challenging task as poems carry a deeper meaning, which can be easily lost if only the literal meaning is considered. That being said, we propose a new task in the field of natural language understanding called {`}Poem Summarization{'}. As a starting, we propose the first-ever dataset for this task, named {`}PoemSum{'}, consisting of 3011 samples of poetry and its corresponding summarized interpretation in the English language. We have benchmarked the performance of different state-of-the-art summarization models and provided observations on their limitations. The dataset and all relevant code used in this work have been made publicly available.",
}
| While research in natural language processing has progressed significantly in creative language generation, the question of whether language models can interpret the intended meaning of creative language largely remains unanswered. Poetry as a creative art form has existed for generations, and summarization of such content requires deciphering the figurative patterns to find out the actual intent and message of the poet. This task can provide the researchers an opportunity to evaluate the creative language interpretation capacity of the language models. Unlike typical text, summarization of poems is a challenging task as poems carry a deeper meaning, which can be easily lost if only the literal meaning is considered. That being said, we propose a new task in the field of natural language understanding called {`}Poem Summarization{'}. As a starting, we propose the first-ever dataset for this task, named {`}PoemSum{'}, consisting of 3011 samples of poetry and its corresponding summarized interpretation in the English language. We have benchmarked the performance of different state-of-the-art summarization models and provided observations on their limitations. The dataset and all relevant code used in this work have been made publicly available. | [
"Mahbub, Ridwan",
"Khan, Ifrad",
"Anuva, Samiha",
"Shahriar, Md Shihab",
"Laskar, Md Tahmid Rahman",
"Ahmed, Sabbir"
] | Unveiling the Essence of Poetry: Introducing a Comprehensive Dataset and Benchmark for Poem Summarization | emnlp-main.920 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Oral |
||
https://aclanthology.org/2023.emnlp-main.921.bib | https://aclanthology.org/2023.emnlp-main.921/ | @inproceedings{huang-etal-2023-privacy,
title = "Privacy Implications of Retrieval-Based Language Models",
author = "Huang, Yangsibo and
Gupta, Samyak and
Zhong, Zexuan and
Li, Kai and
Chen, Danqi",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.921",
doi = "10.18653/v1/2023.emnlp-main.921",
pages = "14887--14902",
abstract = "Retrieval-based language models (LMs) have demonstrated improved interpretability, factuality, and adaptability compared to their parametric counterparts by incorporating retrieved text from external datastores. While it is well known that parametric models are prone to leaking private data, it remains unclear how the addition of a retrieval datastore impacts model privacy. In this work, we present the first study of privacy risks in retrieval-based LMs, particularly $k$NN-LMs. Our goal is to explore the optimal design and training procedure in domains where privacy is of concern, aiming to strike a balance between utility and privacy. Crucially, we find that $k$NN-LMs are more susceptible to leaking private information from their private datastore than parametric models. We further explore mitigations of privacy risks: When privacy information is targeted and readily detected in the text, we find that a simple sanitization step would eliminate the risks while decoupling query and key encoders achieves an even better utility-privacy trade-off. Otherwise, we consider strategies of mixing public and private data in both datastore and encoder training. While these methods offer modest improvements, they leave considerable room for future work. Together, our findings provide insights for practitioners to better understand and mitigate privacy risks in retrieval-based LMs.",
}
| Retrieval-based language models (LMs) have demonstrated improved interpretability, factuality, and adaptability compared to their parametric counterparts by incorporating retrieved text from external datastores. While it is well known that parametric models are prone to leaking private data, it remains unclear how the addition of a retrieval datastore impacts model privacy. In this work, we present the first study of privacy risks in retrieval-based LMs, particularly $k$NN-LMs. Our goal is to explore the optimal design and training procedure in domains where privacy is of concern, aiming to strike a balance between utility and privacy. Crucially, we find that $k$NN-LMs are more susceptible to leaking private information from their private datastore than parametric models. We further explore mitigations of privacy risks: When privacy information is targeted and readily detected in the text, we find that a simple sanitization step would eliminate the risks while decoupling query and key encoders achieves an even better utility-privacy trade-off. Otherwise, we consider strategies of mixing public and private data in both datastore and encoder training. While these methods offer modest improvements, they leave considerable room for future work. Together, our findings provide insights for practitioners to better understand and mitigate privacy risks in retrieval-based LMs. | [
"Huang, Yangsibo",
"Gupta, Samyak",
"Zhong, Zexuan",
"Li, Kai",
"Chen, Danqi"
] | Privacy Implications of Retrieval-Based Language Models | emnlp-main.921 | 2305.14888 | [
"https://github.com/princeton-sysml/knnlm_privacy"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.922.bib | https://aclanthology.org/2023.emnlp-main.922/ | @inproceedings{huang-etal-2023-imtlab,
title = "{IMTL}ab: An Open-Source Platform for Building, Evaluating, and Diagnosing Interactive Machine Translation Systems",
author = "Huang, Xu and
Zhang, Zhirui and
Gao, Ruize and
Du, Yichao and
Liu, Lemao and
Huang, Guoping and
Shi, Shuming and
Chen, Jiajun and
Huang, Shujian",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.922",
doi = "10.18653/v1/2023.emnlp-main.922",
pages = "14903--14917",
abstract = "We present IMTLab, an open-source end-to-end interactive machine translation (IMT) system platform that enables researchers to quickly build IMT systems with state-of-the-art models, perform an end-to-end evaluation, and diagnose the weakness of systems. IMTLab treats the whole interactive translation process as a task-oriented dialogue with a human-in-the-loop setting, in which human interventions can be explicitly incorporated to produce high-quality, error-free translations. To this end, a general communication interface is designed to support the flexible IMT architectures and user policies. Based on the proposed design, we construct a simulated and real interactive environment to achieve end-to-end evaluation and leverage the framework to systematically evaluate previous IMT systems. Our simulated and manual experiments show that the prefix-constrained decoding approach still gains the lowest editing cost in the end-to-end evaluation, while BiTIIMT achieves comparable editing cost with a better interactive experience.",
}
| We present IMTLab, an open-source end-to-end interactive machine translation (IMT) system platform that enables researchers to quickly build IMT systems with state-of-the-art models, perform an end-to-end evaluation, and diagnose the weakness of systems. IMTLab treats the whole interactive translation process as a task-oriented dialogue with a human-in-the-loop setting, in which human interventions can be explicitly incorporated to produce high-quality, error-free translations. To this end, a general communication interface is designed to support the flexible IMT architectures and user policies. Based on the proposed design, we construct a simulated and real interactive environment to achieve end-to-end evaluation and leverage the framework to systematically evaluate previous IMT systems. Our simulated and manual experiments show that the prefix-constrained decoding approach still gains the lowest editing cost in the end-to-end evaluation, while BiTIIMT achieves comparable editing cost with a better interactive experience. | [
"Huang, Xu",
"Zhang, Zhirui",
"Gao, Ruize",
"Du, Yichao",
"Liu, Lemao",
"Huang, Guoping",
"Shi, Shuming",
"Chen, Jiajun",
"Huang, Shujian"
] | IMTLab: An Open-Source Platform for Building, Evaluating, and Diagnosing Interactive Machine Translation Systems | emnlp-main.922 | 2310.11163 | [
"https://github.com/xuuhuang/imtlab"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Oral |
|
https://aclanthology.org/2023.emnlp-main.923.bib | https://aclanthology.org/2023.emnlp-main.923/ | @inproceedings{sun-etal-2023-chatgpt,
title = "Is {C}hat{GPT} Good at Search? Investigating Large Language Models as Re-Ranking Agents",
author = "Sun, Weiwei and
Yan, Lingyong and
Ma, Xinyu and
Wang, Shuaiqiang and
Ren, Pengjie and
Chen, Zhumin and
Yin, Dawei and
Ren, Zhaochun",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.923",
doi = "10.18653/v1/2023.emnlp-main.923",
pages = "14918--14937",
abstract = "Large Language Models (LLMs) have demonstrated remarkable zero-shot generalization across various language-related tasks, including search engines. However, existing work utilizes the generative ability of LLMs for Information Retrieval (IR) rather than direct passage ranking. The discrepancy between the pre-training objectives of LLMs and the ranking objective poses another challenge. In this paper, we first investigate generative LLMs such as ChatGPT and GPT-4 for relevance ranking in IR. Surprisingly, our experiments reveal that properly instructed LLMs can deliver competitive, even superior results to state-of-the-art supervised methods on popular IR benchmarks. Furthermore, to address concerns about data contamination of LLMs, we collect a new test set called NovelEval, based on the latest knowledge and aiming to verify the model{'}s ability to rank unknown knowledge. Finally, to improve efficiency in real-world applications, we delve into the potential for distilling the ranking capabilities of ChatGPT into small specialized models using a permutation distillation scheme. Our evaluation results turn out that a distilled 440M model outperforms a 3B supervised model on the BEIR benchmark. The code to reproduce our results is available at www.github.com/sunnweiwei/RankGPT.",
}
| Large Language Models (LLMs) have demonstrated remarkable zero-shot generalization across various language-related tasks, including search engines. However, existing work utilizes the generative ability of LLMs for Information Retrieval (IR) rather than direct passage ranking. The discrepancy between the pre-training objectives of LLMs and the ranking objective poses another challenge. In this paper, we first investigate generative LLMs such as ChatGPT and GPT-4 for relevance ranking in IR. Surprisingly, our experiments reveal that properly instructed LLMs can deliver competitive, even superior results to state-of-the-art supervised methods on popular IR benchmarks. Furthermore, to address concerns about data contamination of LLMs, we collect a new test set called NovelEval, based on the latest knowledge and aiming to verify the model{'}s ability to rank unknown knowledge. Finally, to improve efficiency in real-world applications, we delve into the potential for distilling the ranking capabilities of ChatGPT into small specialized models using a permutation distillation scheme. Our evaluation results turn out that a distilled 440M model outperforms a 3B supervised model on the BEIR benchmark. The code to reproduce our results is available at www.github.com/sunnweiwei/RankGPT. | [
"Sun, Weiwei",
"Yan, Lingyong",
"Ma, Xinyu",
"Wang, Shuaiqiang",
"Ren, Pengjie",
"Chen, Zhumin",
"Yin, Dawei",
"Ren, Zhaochun"
] | Is ChatGPT Good at Search? Investigating Large Language Models as Re-Ranking Agents | emnlp-main.923 | 2304.09542 | [
"https://github.com/sunnweiwei/rankgpt"
] | https://huggingface.co/papers/2304.09542 | 0 | 4 | 0 | 6 | [
"princeton-nlp/Llama-3-8B-ProLong-64k-Instruct",
"QuantFactory/Llama-3-8B-ProLong-64k-Instruct-GGUF",
"princeton-nlp/Llama-3-8B-ProLong-64k-Base",
"RichardErkhov/princeton-nlp_-_Llama-3-8B-ProLong-64k-Base-gguf"
] | [] | [] | 1 | Oral |
https://aclanthology.org/2023.emnlp-main.924.bib | https://aclanthology.org/2023.emnlp-main.924/ | @inproceedings{hu-etal-2023-diner,
title = "{D}i{N}e{R}: A Large Realistic Dataset for Evaluating Compositional Generalization",
author = "Hu, Chengang and
Liu, Xiao and
Feng, Yansong",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.924",
doi = "10.18653/v1/2023.emnlp-main.924",
pages = "14938--14947",
abstract = "Most of the existing compositional generalization datasets are synthetically-generated, resulting in a lack of natural language variation. While there have been recent attempts to introduce non-synthetic datasets for compositional generalization, they suffer from either limited data scale or a lack of diversity in the forms of combinations. To better investigate compositional generalization with more linguistic phenomena and compositional diversity, we propose the DIsh NamE Recognition (DiNeR) task and create a large realistic Chinese dataset. Given a recipe instruction, models are required to recognize the dish name composed of diverse combinations of food, actions, and flavors. Our dataset consists of 3,811 dishes and 228,114 recipes, and involves plenty of linguistic phenomena such as anaphora, omission and ambiguity. We provide two strong baselines based on T5 and large language models (LLMs). This work contributes a challenging task, baseline methods to tackle the task, and insights into compositional generalization in the context of dish name recognition.",
}
| Most of the existing compositional generalization datasets are synthetically-generated, resulting in a lack of natural language variation. While there have been recent attempts to introduce non-synthetic datasets for compositional generalization, they suffer from either limited data scale or a lack of diversity in the forms of combinations. To better investigate compositional generalization with more linguistic phenomena and compositional diversity, we propose the DIsh NamE Recognition (DiNeR) task and create a large realistic Chinese dataset. Given a recipe instruction, models are required to recognize the dish name composed of diverse combinations of food, actions, and flavors. Our dataset consists of 3,811 dishes and 228,114 recipes, and involves plenty of linguistic phenomena such as anaphora, omission and ambiguity. We provide two strong baselines based on T5 and large language models (LLMs). This work contributes a challenging task, baseline methods to tackle the task, and insights into compositional generalization in the context of dish name recognition. | [
"Hu, Chengang",
"Liu, Xiao",
"Feng, Yansong"
] | DiNeR: A Large Realistic Dataset for Evaluating Compositional Generalization | emnlp-main.924 | 2406.04669 | [
"https://github.com/jumpy-pku/diner"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.925.bib | https://aclanthology.org/2023.emnlp-main.925/ | @inproceedings{chen-etal-2023-pre-trained,
title = "Can Pre-trained Vision and Language Models Answer Visual Information-Seeking Questions?",
author = "Chen, Yang and
Hu, Hexiang and
Luan, Yi and
Sun, Haitian and
Changpinyo, Soravit and
Ritter, Alan and
Chang, Ming-Wei",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.925",
doi = "10.18653/v1/2023.emnlp-main.925",
pages = "14948--14968",
abstract = "Pre-trained vision and language models have demonstrated state-of-the-art capabilities over existing tasks involving images and texts, including visual question answering. However, it remains unclear whether these models possess the capability to answer questions that are not only querying visual content but knowledge-intensive and information-seeking. In this study, we introduce InfoSeek, a visual question answering dataset tailored for information-seeking questions that cannot be answered with only common sense knowledge. Using InfoSeek, we analyze various pre-trained visual question answering models and gain insights into their characteristics. Our findings reveal that state-of-the-art pre-trained multi-modal models (e.g., PaLI-X, BLIP2, InstructBLIP) face challenges in answering visual information-seeking questions, but fine-tuning on the InfoSeek dataset elicits models to use fine-grained knowledge that was learned during pre-training. Furthermore, we show that accurate visual entity recognition can be used to improve performance on InfoSeek by retrieving relevant documents, showing a significant space for improvement.",
}
| Pre-trained vision and language models have demonstrated state-of-the-art capabilities over existing tasks involving images and texts, including visual question answering. However, it remains unclear whether these models possess the capability to answer questions that are not only querying visual content but knowledge-intensive and information-seeking. In this study, we introduce InfoSeek, a visual question answering dataset tailored for information-seeking questions that cannot be answered with only common sense knowledge. Using InfoSeek, we analyze various pre-trained visual question answering models and gain insights into their characteristics. Our findings reveal that state-of-the-art pre-trained multi-modal models (e.g., PaLI-X, BLIP2, InstructBLIP) face challenges in answering visual information-seeking questions, but fine-tuning on the InfoSeek dataset elicits models to use fine-grained knowledge that was learned during pre-training. Furthermore, we show that accurate visual entity recognition can be used to improve performance on InfoSeek by retrieving relevant documents, showing a significant space for improvement. | [
"Chen, Yang",
"Hu, Hexiang",
"Luan, Yi",
"Sun, Haitian",
"Changpinyo, Soravit",
"Ritter, Alan",
"Chang, Ming-Wei"
] | Can Pre-trained Vision and Language Models Answer Visual Information-Seeking Questions? | emnlp-main.925 | 2302.11713 | [
"https://github.com/edchengg/infoseek_eval"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.926.bib | https://aclanthology.org/2023.emnlp-main.926/ | @inproceedings{li-etal-2023-eder,
title = "{ED}e{R}: Towards Understanding Dependency Relations Between Events",
author = "Li, Ruiqi and
Haslum, Patrik and
Cui, Leyang",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.926",
doi = "10.18653/v1/2023.emnlp-main.926",
pages = "14969--14983",
abstract = "Relation extraction is a crucial task in natural language processing (NLP) and information retrieval (IR). Previous work on event relation extraction mainly focuses on hierarchical, temporal and causal relations. Such relationships consider two events to be independent in terms of syntax and semantics, but they fail to recognize the interdependence between events. To bridge this gap, we introduce a human-annotated Event Dependency Relation dataset (EDeR). The annotation is done on a sample of documents from the OntoNotes dataset, which has the additional benefit that it integrates with existing, orthogonal, annotations of this dataset. We investigate baseline approaches for EDeR{'}s event dependency relation prediction. We show that recognizing such event dependency relations can further benefit critical NLP tasks, including semantic role labelling and co-reference resolution.",
}
| Relation extraction is a crucial task in natural language processing (NLP) and information retrieval (IR). Previous work on event relation extraction mainly focuses on hierarchical, temporal and causal relations. Such relationships consider two events to be independent in terms of syntax and semantics, but they fail to recognize the interdependence between events. To bridge this gap, we introduce a human-annotated Event Dependency Relation dataset (EDeR). The annotation is done on a sample of documents from the OntoNotes dataset, which has the additional benefit that it integrates with existing, orthogonal, annotations of this dataset. We investigate baseline approaches for EDeR{'}s event dependency relation prediction. We show that recognizing such event dependency relations can further benefit critical NLP tasks, including semantic role labelling and co-reference resolution. | [
"Li, Ruiqi",
"Haslum, Patrik",
"Cui, Leyang"
] | EDeR: Towards Understanding Dependency Relations Between Events | emnlp-main.926 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
||
https://aclanthology.org/2023.emnlp-main.927.bib | https://aclanthology.org/2023.emnlp-main.927/ | @inproceedings{kim-etal-2023-aint,
title = "It Ain{'}t Over: A Multi-aspect Diverse Math Word Problem Dataset",
author = "Kim, Jiwoo and
Kim, Youngbin and
Baek, Ilwoong and
Bak, JinYeong and
Lee, Jongwuk",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.927",
doi = "10.18653/v1/2023.emnlp-main.927",
pages = "14984--15011",
abstract = "The math word problem (MWP) is a complex task that requires natural language understanding and logical reasoning to extract key knowledge from natural language narratives. Previous studies have provided various MWP datasets but lack diversity in problem types, lexical usage patterns, languages, and annotations for intermediate solutions. To address these limitations, we introduce a new MWP dataset, named DMath (Diverse Math Word Problems), offering a wide range of diversity in problem types, lexical usage patterns, languages, and intermediate solutions. The problems are available in English and Korean and include an expression tree and Python code as intermediate solutions. Through extensive experiments, we demonstrate that the DMath dataset provides a new opportunity to evaluate the capability of large language models, i.e., GPT-4 only achieves about 75{\%} accuracy on the DMath dataset.",
}
| The math word problem (MWP) is a complex task that requires natural language understanding and logical reasoning to extract key knowledge from natural language narratives. Previous studies have provided various MWP datasets but lack diversity in problem types, lexical usage patterns, languages, and annotations for intermediate solutions. To address these limitations, we introduce a new MWP dataset, named DMath (Diverse Math Word Problems), offering a wide range of diversity in problem types, lexical usage patterns, languages, and intermediate solutions. The problems are available in English and Korean and include an expression tree and Python code as intermediate solutions. Through extensive experiments, we demonstrate that the DMath dataset provides a new opportunity to evaluate the capability of large language models, i.e., GPT-4 only achieves about 75{\%} accuracy on the DMath dataset. | [
"Kim, Jiwoo",
"Kim, Youngbin",
"Baek, Ilwoong",
"Bak, JinYeong",
"Lee, Jongwuk"
] | It Ain't Over: A Multi-aspect Diverse Math Word Problem Dataset | emnlp-main.927 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Oral |
||
https://aclanthology.org/2023.emnlp-main.928.bib | https://aclanthology.org/2023.emnlp-main.928/ | @inproceedings{koopman-zuccon-2023-dr,
title = "Dr {C}hat{GPT} tell me what {I} want to hear: How different prompts impact health answer correctness",
author = "Koopman, Bevan and
Zuccon, Guido",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.928",
doi = "10.18653/v1/2023.emnlp-main.928",
pages = "15012--15022",
abstract = "This paper investigates the significant impact different prompts have on the behaviour of ChatGPT when used for health information seeking. As people more and more depend on generative large language models (LLMs) like ChatGPT, it is critical to understand model behaviour under different conditions, especially for domains where incorrect answers can have serious consequences such as health. Using the TREC Misinformation dataset, we empirically evaluate ChatGPT to show not just its effectiveness but reveal that knowledge passed in the prompt can bias the model to the detriment of answer correctness. We show this occurs both for retrieve-then-generate pipelines and based on how a user phrases their question as well as the question type. This work has important implications for the development of more robust and transparent question-answering systems based on generative large language models. Prompts, raw result files and manual analysis are made publicly available at \url{https://github.com/ielab/drchatgpt-health_prompting}.",
}
| This paper investigates the significant impact different prompts have on the behaviour of ChatGPT when used for health information seeking. As people more and more depend on generative large language models (LLMs) like ChatGPT, it is critical to understand model behaviour under different conditions, especially for domains where incorrect answers can have serious consequences such as health. Using the TREC Misinformation dataset, we empirically evaluate ChatGPT to show not just its effectiveness but reveal that knowledge passed in the prompt can bias the model to the detriment of answer correctness. We show this occurs both for retrieve-then-generate pipelines and based on how a user phrases their question as well as the question type. This work has important implications for the development of more robust and transparent question-answering systems based on generative large language models. Prompts, raw result files and manual analysis are made publicly available at \url{https://github.com/ielab/drchatgpt-health_prompting}. | [
"Koopman, Bevan",
"Zuccon, Guido"
] | Dr ChatGPT tell me what I want to hear: How different prompts impact health answer correctness | emnlp-main.928 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Oral |
||
https://aclanthology.org/2023.emnlp-main.929.bib | https://aclanthology.org/2023.emnlp-main.929/ | @inproceedings{wang-etal-2023-knn,
title = "$k${NN}-{LM} Does Not Improve Open-ended Text Generation",
author = "Wang, Shufan and
Song, Yixiao and
Drozdov, Andrew and
Garimella, Aparna and
Manjunatha, Varun and
Iyyer, Mohit",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.929",
doi = "10.18653/v1/2023.emnlp-main.929",
pages = "15023--15037",
abstract = "In this paper, we study the generation quality of interpolation-based retrieval-augmented language models (LMs). These methods, best exemplified by the $k$NN-LM, interpolate the LM{'}s predicted distribution of the next word with a distribution formed from the most relevant retrievals for a given prefix. While the $k$NN-LM and related methods yield impressive decreases in perplexity, we discover that they do not exhibit corresponding improvements in open-ended generation quality, as measured by both automatic evaluation metrics (e.g., MAUVE) and human evaluations. Digging deeper, we find that interpolating with a retrieval distribution actually increases perplexity compared to a baseline LM for the majority of tokens in the WikiText-103 test set, even though the overall perplexity is lower due to a smaller number of tokens for which perplexity dramatically decreases after interpolation. However, when decoding a long sequence at inference time, significant improvements on this smaller subset of tokens are washed out by slightly worse predictions on most tokens. Furthermore, we discover that the entropy of the retrieval distribution increases faster than that of the base LM as the generated sequence becomes longer, which indicates that retrieval is less reliable when using model-generated text as queries (i.e., is subject to exposure bias). We hope that our analysis spurs future work on improved decoding algorithms and interpolation strategies for retrieval-augmented language models.",
}
| In this paper, we study the generation quality of interpolation-based retrieval-augmented language models (LMs). These methods, best exemplified by the $k$NN-LM, interpolate the LM{'}s predicted distribution of the next word with a distribution formed from the most relevant retrievals for a given prefix. While the $k$NN-LM and related methods yield impressive decreases in perplexity, we discover that they do not exhibit corresponding improvements in open-ended generation quality, as measured by both automatic evaluation metrics (e.g., MAUVE) and human evaluations. Digging deeper, we find that interpolating with a retrieval distribution actually increases perplexity compared to a baseline LM for the majority of tokens in the WikiText-103 test set, even though the overall perplexity is lower due to a smaller number of tokens for which perplexity dramatically decreases after interpolation. However, when decoding a long sequence at inference time, significant improvements on this smaller subset of tokens are washed out by slightly worse predictions on most tokens. Furthermore, we discover that the entropy of the retrieval distribution increases faster than that of the base LM as the generated sequence becomes longer, which indicates that retrieval is less reliable when using model-generated text as queries (i.e., is subject to exposure bias). We hope that our analysis spurs future work on improved decoding algorithms and interpolation strategies for retrieval-augmented language models. | [
"Wang, Shufan",
"Song, Yixiao",
"Drozdov, Andrew",
"Garimella, Aparna",
"Manjunatha, Varun",
"Iyyer, Mohit"
] | kNN-LM Does Not Improve Open-ended Text Generation | emnlp-main.929 | 2305.14625 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.930.bib | https://aclanthology.org/2023.emnlp-main.930/ | @inproceedings{liu-etal-2023-towards-unified,
title = "Towards A Unified View of Sparse Feed-Forward Network in Pretraining Large Language Model",
author = "Liu, Zeyu and
Dettmers, Tim and
Lin, Xi and
Stoyanov, Veselin and
Li, Xian",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.930",
doi = "10.18653/v1/2023.emnlp-main.930",
pages = "15038--15061",
abstract = "Large and sparse feed-forward layers (S-FFN) such as Mixture-of-Experts (MoE) have proven effective in scaling up Transformers model size for pretraining large language models. By only activating part of the FFN parameters conditioning on input, S-FFN improves generalization performance while keeping training and inference costs (in FLOPs) fixed. In this work, we analyzed two major design choices of S-FFN: the memory block (a.k.a. expert) size and the memory block selection method under a general conceptual framework of sparse neural memory. Using this unified framework, we compare several S-FFN architectures for language modeling and provide insights into their relative efficacy and efficiency. We found a simpler selection method {---} Avg-K that selects blocks through their mean aggregated hidden states, achieving lower perplexity in language model pretraining compared to existing MoE architectures including Switch Transformer (Fedus et al., 2021) and HashLayer (Roller et al., 2021).",
}
| Large and sparse feed-forward layers (S-FFN) such as Mixture-of-Experts (MoE) have proven effective in scaling up Transformers model size for pretraining large language models. By only activating part of the FFN parameters conditioning on input, S-FFN improves generalization performance while keeping training and inference costs (in FLOPs) fixed. In this work, we analyzed two major design choices of S-FFN: the memory block (a.k.a. expert) size and the memory block selection method under a general conceptual framework of sparse neural memory. Using this unified framework, we compare several S-FFN architectures for language modeling and provide insights into their relative efficacy and efficiency. We found a simpler selection method {---} Avg-K that selects blocks through their mean aggregated hidden states, achieving lower perplexity in language model pretraining compared to existing MoE architectures including Switch Transformer (Fedus et al., 2021) and HashLayer (Roller et al., 2021). | [
"Liu, Zeyu",
"Dettmers, Tim",
"Lin, Xi",
"Stoyanov, Veselin",
"Li, Xian"
] | Towards A Unified View of Sparse Feed-Forward Network in Pretraining Large Language Model | emnlp-main.930 | 2305.13999 | [
""
] | https://huggingface.co/papers/2305.13999 | 0 | 0 | 0 | 5 | [] | [] | [] | 1 | Poster |
https://aclanthology.org/2023.emnlp-main.931.bib | https://aclanthology.org/2023.emnlp-main.931/ | @inproceedings{su-etal-2023-exploring,
title = "Exploring the Impact of Model Scaling on Parameter-Efficient Tuning",
author = "Su, Yusheng and
Chan, Chi-Min and
Cheng, Jiali and
Qin, Yujia and
Lin, Yankai and
Hu, Shengding and
Yang, Zonghan and
Ding, Ning and
Sun, Xingzhi and
Xie, Guotong and
Liu, Zhiyuan and
Sun, Maosong",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.931",
doi = "10.18653/v1/2023.emnlp-main.931",
pages = "15062--15078",
abstract = "Parameter-efficient tuning (PET) methods can effectively drive extremely large pre-trained language models (PLMs) by training only minimal parameters. Different PET methods utilize different manually designed tunable modules. In small PLMs, there are usually noticeable performance differences among PET methods. Nevertheless, as the model scale increases, the performance differences become marginal. Hence, we hypothesize that model scaling mitigates the impact of design differences on PET methods. To investigate this hypothesis, we introduce a more flexible PET method called Arbitrary PET (APET) method. The APET method is compatible with a tunable module, which consists of any number of parameters distributed in arbitrary positions. Then, we utilize it and conduct experiments on 11 NLP tasks across 3 representative PLMs. Our investigations reveal that model scaling (1) mitigates the effects of the positions of tunable parameters on performance, and (2) enables tuning methods to achieve performance comparable to full-parameter fine-tuning by optimizing fewer tunable parameters. Intriguingly, we also observe that tuning methods optimize the similar number of tunable parameters to exceed random guess performance on different tasks. We collectively discuss this phenomenon and the two aforementioned findings from an optimization perspective to understand the underlying mechanisms. These conclusions enhance our understanding of the impact of model scaling on PET and assist in designing more effective and efficient PET methods for PLMs of different scales. The source code can be obtained from this GitHub repository: \url{https://github.com/yushengsu-thu/PET_Scaling}.",
}
| Parameter-efficient tuning (PET) methods can effectively drive extremely large pre-trained language models (PLMs) by training only minimal parameters. Different PET methods utilize different manually designed tunable modules. In small PLMs, there are usually noticeable performance differences among PET methods. Nevertheless, as the model scale increases, the performance differences become marginal. Hence, we hypothesize that model scaling mitigates the impact of design differences on PET methods. To investigate this hypothesis, we introduce a more flexible PET method called Arbitrary PET (APET) method. The APET method is compatible with a tunable module, which consists of any number of parameters distributed in arbitrary positions. Then, we utilize it and conduct experiments on 11 NLP tasks across 3 representative PLMs. Our investigations reveal that model scaling (1) mitigates the effects of the positions of tunable parameters on performance, and (2) enables tuning methods to achieve performance comparable to full-parameter fine-tuning by optimizing fewer tunable parameters. Intriguingly, we also observe that tuning methods optimize the similar number of tunable parameters to exceed random guess performance on different tasks. We collectively discuss this phenomenon and the two aforementioned findings from an optimization perspective to understand the underlying mechanisms. These conclusions enhance our understanding of the impact of model scaling on PET and assist in designing more effective and efficient PET methods for PLMs of different scales. The source code can be obtained from this GitHub repository: \url{https://github.com/yushengsu-thu/PET_Scaling}. | [
"Su, Yusheng",
"Chan, Chi-Min",
"Cheng, Jiali",
"Qin, Yujia",
"Lin, Yankai",
"Hu, Shengding",
"Yang, Zonghan",
"Ding, Ning",
"Sun, Xingzhi",
"Xie, Guotong",
"Liu, Zhiyuan",
"Sun, Maosong"
] | Exploring the Impact of Model Scaling on Parameter-Efficient Tuning | emnlp-main.931 | 2306.02320 | [
"https://github.com/yushengsu-thu/pet_scaling"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.932.bib | https://aclanthology.org/2023.emnlp-main.932/ | @inproceedings{chen-etal-2023-stair,
title = "{STAIR}: Learning Sparse Text and Image Representation in Grounded Tokens",
author = "Chen, Chen and
Zhang, Bowen and
Cao, Liangliang and
Shen, Jiguang and
Gunter, Tom and
Jose, Albin and
Toshev, Alexander and
Zheng, Yantao and
Shlens, Jonathon and
Pang, Ruoming and
Yang, Yinfei",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.932",
doi = "10.18653/v1/2023.emnlp-main.932",
pages = "15079--15094",
abstract = "Image and text retrieval is one of the foundational tasks in the vision and language domain with multiple real-world applications. State-of-the-art contrastive approaches, e.g. CLIP, ALIGN, represent images and texts as dense embeddings and calculate the similarity in the dense embedding space as the matching score. On the other hand, sparse semantic features like bag-of-words models are more interpretable, but believed to suffer from inferior accuracy than dense representations. In this work, we show that it is possible to build a sparse semantic representation that is as powerful as, or even better than, dense presentations. We extend the CLIP model and build a sparse text and image representation (STAIR), where the image and text are mapped to a sparse token space. Each token in the space is a (sub-)word in the vocabulary, which is not only interpretable but also easy to integrate with existing information retrieval systems. STAIR model significantly outperforms a CLIP model with +4.9{\%} and +4.3{\%} absolute Recall@1 improvement on COCO-5k text$\rightarrow$image and image$\rightarrow$text retrieval respectively. It also achieved better performance on both of ImageNet zero-shot and linear probing compared to CLIP.",
}
| Image and text retrieval is one of the foundational tasks in the vision and language domain with multiple real-world applications. State-of-the-art contrastive approaches, e.g. CLIP, ALIGN, represent images and texts as dense embeddings and calculate the similarity in the dense embedding space as the matching score. On the other hand, sparse semantic features like bag-of-words models are more interpretable, but believed to suffer from inferior accuracy than dense representations. In this work, we show that it is possible to build a sparse semantic representation that is as powerful as, or even better than, dense presentations. We extend the CLIP model and build a sparse text and image representation (STAIR), where the image and text are mapped to a sparse token space. Each token in the space is a (sub-)word in the vocabulary, which is not only interpretable but also easy to integrate with existing information retrieval systems. STAIR model significantly outperforms a CLIP model with +4.9{\%} and +4.3{\%} absolute Recall@1 improvement on COCO-5k text$\rightarrow$image and image$\rightarrow$text retrieval respectively. It also achieved better performance on both of ImageNet zero-shot and linear probing compared to CLIP. | [
"Chen, Chen",
"Zhang, Bowen",
"Cao, Liangliang",
"Shen, Jiguang",
"Gunter, Tom",
"Jose, Albin",
"Toshev, Alex",
"er",
"Zheng, Yantao",
"Shlens, Jonathon",
"Pang, Ruoming",
"Yang, Yinfei"
] | STAIR: Learning Sparse Text and Image Representation in Grounded Tokens | emnlp-main.932 | 2301.13081 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.933.bib | https://aclanthology.org/2023.emnlp-main.933/ | @inproceedings{liu-etal-2023-crossing,
title = "Crossing the Threshold: Idiomatic Machine Translation through Retrieval Augmentation and Loss Weighting",
author = "Liu, Emmy and
Chaudhary, Aditi and
Neubig, Graham",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.933",
doi = "10.18653/v1/2023.emnlp-main.933",
pages = "15095--15111",
abstract = "Idioms are common in everyday language, but often pose a challenge to translators because their meanings do not follow from the meanings of their parts. Despite significant advances, machine translation systems still struggle to translate idiomatic expressions. We provide a simple characterization of idiomatic translation and related issues. This allows us to conduct a synthetic experiment revealing a tipping point at which transformer-based machine translation models correctly default to idiomatic translations. To expand multilingual resources, we compile a dataset of {\textasciitilde}4k natural sentences containing idiomatic expressions in French, Finnish, and Japanese. To improve translation of natural idioms, we introduce two straightforward yet effective techniques: the strategic upweighting of training loss on potentially idiomatic sentences, and using retrieval-augmented models. This not only improves the accuracy of a strong pretrained MT model on idiomatic sentences by up to 13{\%} in absolute accuracy, but also holds potential benefits for non-idiomatic sentences.",
}
| Idioms are common in everyday language, but often pose a challenge to translators because their meanings do not follow from the meanings of their parts. Despite significant advances, machine translation systems still struggle to translate idiomatic expressions. We provide a simple characterization of idiomatic translation and related issues. This allows us to conduct a synthetic experiment revealing a tipping point at which transformer-based machine translation models correctly default to idiomatic translations. To expand multilingual resources, we compile a dataset of {\textasciitilde}4k natural sentences containing idiomatic expressions in French, Finnish, and Japanese. To improve translation of natural idioms, we introduce two straightforward yet effective techniques: the strategic upweighting of training loss on potentially idiomatic sentences, and using retrieval-augmented models. This not only improves the accuracy of a strong pretrained MT model on idiomatic sentences by up to 13{\%} in absolute accuracy, but also holds potential benefits for non-idiomatic sentences. | [
"Liu, Emmy",
"Chaudhary, Aditi",
"Neubig, Graham"
] | Crossing the Threshold: Idiomatic Machine Translation through Retrieval Augmentation and Loss Weighting | emnlp-main.933 | 2310.07081 | [
"https://github.com/nightingal3/idiom-translation"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Oral |
|
https://aclanthology.org/2023.emnlp-main.934.bib | https://aclanthology.org/2023.emnlp-main.934/ | @inproceedings{wang-etal-2023-corec,
title = "{C}o{R}ec: An Easy Approach for Coordination Recognition",
author = "Wang, Qing and
Jia, Haojie and
Song, Wenfei and
Li, Qi",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.934",
doi = "10.18653/v1/2023.emnlp-main.934",
pages = "15112--15120",
abstract = "In this paper, we observe and address the challenges of the coordination recognition task. Most existing methods rely on syntactic parsers to identify the coordinators in a sentence and detect the coordination boundaries. However, state-of-the-art syntactic parsers are slow and suffer from errors, especially for long and complicated sentences. To better solve the problems, we propose a pipeline model COordination RECognizer (CoRec). It consists of two components: coordinator identifier and conjunct boundary detector. The experimental results on datasets from various domains demonstrate the effectiveness and efficiency of the proposed method. Further experiments show that CoRec positively impacts downstream tasks, improving the yield of state-of-the-art Open IE models.",
}
| In this paper, we observe and address the challenges of the coordination recognition task. Most existing methods rely on syntactic parsers to identify the coordinators in a sentence and detect the coordination boundaries. However, state-of-the-art syntactic parsers are slow and suffer from errors, especially for long and complicated sentences. To better solve the problems, we propose a pipeline model COordination RECognizer (CoRec). It consists of two components: coordinator identifier and conjunct boundary detector. The experimental results on datasets from various domains demonstrate the effectiveness and efficiency of the proposed method. Further experiments show that CoRec positively impacts downstream tasks, improving the yield of state-of-the-art Open IE models. | [
"Wang, Qing",
"Jia, Haojie",
"Song, Wenfei",
"Li, Qi"
] | CoRec: An Easy Approach for Coordination Recognition | emnlp-main.934 | 2311.18712 | [
"https://github.com/qingwang-isu/corec"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Oral |
|
https://aclanthology.org/2023.emnlp-main.935.bib | https://aclanthology.org/2023.emnlp-main.935/ | @inproceedings{otao-yamada-2023-linear,
title = "A linear time approximation of {W}asserstein distance with word embedding selection",
author = "Otao, Sho and
Yamada, Makoto",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.935",
doi = "10.18653/v1/2023.emnlp-main.935",
pages = "15121--15134",
abstract = "Wasserstein distance, which can be computed by solving the optimal transport problem, is a powerful method for measuring the dissimilarity between documents. In the NLP community, it is referred to as word mover{'}s distance (WMD). One of the key challenges of Wasserstein distance is its computational cost since it needs cubic time. Although the Sinkhorn algorithm is a powerful tool to speed up to compute the Wasserstein distance, it still requires square time. Recently, a linear time approximation of the Wasserstein distance including the sliced Wasserstein and the tree-Wasserstein distance (TWD) has been proposed. However, a linear time approximation method suffers when the dimensionality of word vectors is high. In this study, we propose a method to combine feature selection and tree approximation of Wasserstein distance to handle high-dimensional problems. More specifically, we use multiple word embeddings and automatically select useful word embeddings in a tree approximation of Wasserstein distance. To this end, we approximate Wasserstein distance for each word vector by tree approximation technique, and select the discriminative (i.e., large Wasserstein distance) word embeddings by solving an entropic regularized maximization problem. Through our experiments on document classification, our proposed method achieved high performance.",
}
| Wasserstein distance, which can be computed by solving the optimal transport problem, is a powerful method for measuring the dissimilarity between documents. In the NLP community, it is referred to as word mover{'}s distance (WMD). One of the key challenges of Wasserstein distance is its computational cost since it needs cubic time. Although the Sinkhorn algorithm is a powerful tool to speed up to compute the Wasserstein distance, it still requires square time. Recently, a linear time approximation of the Wasserstein distance including the sliced Wasserstein and the tree-Wasserstein distance (TWD) has been proposed. However, a linear time approximation method suffers when the dimensionality of word vectors is high. In this study, we propose a method to combine feature selection and tree approximation of Wasserstein distance to handle high-dimensional problems. More specifically, we use multiple word embeddings and automatically select useful word embeddings in a tree approximation of Wasserstein distance. To this end, we approximate Wasserstein distance for each word vector by tree approximation technique, and select the discriminative (i.e., large Wasserstein distance) word embeddings by solving an entropic regularized maximization problem. Through our experiments on document classification, our proposed method achieved high performance. | [
"Otao, Sho",
"Yamada, Makoto"
] | A linear time approximation of Wasserstein distance with word embedding selection | emnlp-main.935 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
||
https://aclanthology.org/2023.emnlp-main.936.bib | https://aclanthology.org/2023.emnlp-main.936/ | @inproceedings{yin-etal-2023-exchange,
title = "Exchange-of-Thought: Enhancing Large Language Model Capabilities through Cross-Model Communication",
author = "Yin, Zhangyue and
Sun, Qiushi and
Chang, Cheng and
Guo, Qipeng and
Dai, Junqi and
Huang, Xuanjing and
Qiu, Xipeng",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.936",
doi = "10.18653/v1/2023.emnlp-main.936",
pages = "15135--15153",
abstract = "Large Language Models (LLMs) have recently made significant strides in complex reasoning tasks through the Chain-of-Thought technique. Despite this progress, their reasoning is often constrained by their intrinsic understanding, lacking external insights. To address this, we propose Exchange-of-Thought (EoT), a novel framework that enables cross-model communication during problem-solving. Drawing inspiration from network topology, EoT integrates four unique communication paradigms: Memory, Report, Relay, and Debate. This paper delves into the communication dynamics and volume associated with each paradigm. To counterbalance the risks of incorrect reasoning chains, we implement a robust confidence evaluation mechanism within these communications. Our experiments across diverse complex reasoning tasks demonstrate that EoT significantly surpasses established baselines, underscoring the value of external insights in enhancing LLM performance. Furthermore, we show that EoT achieves these superior results in a cost-effective manner, marking a promising advancement for efficient and collaborative AI problem-solving.",
}
| Large Language Models (LLMs) have recently made significant strides in complex reasoning tasks through the Chain-of-Thought technique. Despite this progress, their reasoning is often constrained by their intrinsic understanding, lacking external insights. To address this, we propose Exchange-of-Thought (EoT), a novel framework that enables cross-model communication during problem-solving. Drawing inspiration from network topology, EoT integrates four unique communication paradigms: Memory, Report, Relay, and Debate. This paper delves into the communication dynamics and volume associated with each paradigm. To counterbalance the risks of incorrect reasoning chains, we implement a robust confidence evaluation mechanism within these communications. Our experiments across diverse complex reasoning tasks demonstrate that EoT significantly surpasses established baselines, underscoring the value of external insights in enhancing LLM performance. Furthermore, we show that EoT achieves these superior results in a cost-effective manner, marking a promising advancement for efficient and collaborative AI problem-solving. | [
"Yin, Zhangyue",
"Sun, Qiushi",
"Chang, Cheng",
"Guo, Qipeng",
"Dai, Junqi",
"Huang, Xuanjing",
"Qiu, Xipeng"
] | Exchange-of-Thought: Enhancing Large Language Model Capabilities through Cross-Model Communication | emnlp-main.936 | 2312.01823 | [
"https://github.com/yinzhangyue/eot"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.937.bib | https://aclanthology.org/2023.emnlp-main.937/ | @inproceedings{nguyen-etal-2023-conversation,
title = "Conversation Understanding using Relational Temporal Graph Neural Networks with Auxiliary Cross-Modality Interaction",
author = "Nguyen, Cam Van Thi and
Mai, Tuan and
The, Son and
Kieu, Dang and
Le, Duc-Trong",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.937",
doi = "10.18653/v1/2023.emnlp-main.937",
pages = "15154--15167",
abstract = "Emotion recognition is a crucial task for human conversation understanding. It becomes more challenging with the notion of multimodal data, e.g., language, voice, and facial expressions. As a typical solution, the global- and the local context information are exploited to predict the emotional label for every single sentence, i.e., utterance, in the dialogue. Specifically, the global representation could be captured via modeling of cross-modal interactions at the conversation level. The local one is often inferred using the temporal information of speakers or emotional shifts, which neglects vital factors at the utterance level. Additionally, most existing approaches take fused features of multiple modalities in an unified input without leveraging modality-specific representations. Motivating from these problems, we propose the Relational Temporal Graph Neural Network with Auxiliary Cross-Modality Interaction (CORECT), an novel neural network framework that effectively captures conversation-level cross-modality interactions and utterance-level temporal dependencies with the modality-specific manner for conversation understanding. Extensive experiments demonstrate the effectiveness of CORECT via its state-of-the-art results on the IEMOCAP and CMU-MOSEI datasets for the multimodal ERC task.",
}
| Emotion recognition is a crucial task for human conversation understanding. It becomes more challenging with the notion of multimodal data, e.g., language, voice, and facial expressions. As a typical solution, the global- and the local context information are exploited to predict the emotional label for every single sentence, i.e., utterance, in the dialogue. Specifically, the global representation could be captured via modeling of cross-modal interactions at the conversation level. The local one is often inferred using the temporal information of speakers or emotional shifts, which neglects vital factors at the utterance level. Additionally, most existing approaches take fused features of multiple modalities in an unified input without leveraging modality-specific representations. Motivating from these problems, we propose the Relational Temporal Graph Neural Network with Auxiliary Cross-Modality Interaction (CORECT), an novel neural network framework that effectively captures conversation-level cross-modality interactions and utterance-level temporal dependencies with the modality-specific manner for conversation understanding. Extensive experiments demonstrate the effectiveness of CORECT via its state-of-the-art results on the IEMOCAP and CMU-MOSEI datasets for the multimodal ERC task. | [
"Nguyen, Cam Van Thi",
"Mai, Tuan",
"The, Son",
"Kieu, Dang",
"Le, Duc-Trong"
] | Conversation Understanding using Relational Temporal Graph Neural Networks with Auxiliary Cross-Modality Interaction | emnlp-main.937 | 2311.04507 | [
"https://github.com/leson502/CORECT_EMNLP2023"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Oral |
|
https://aclanthology.org/2023.emnlp-main.938.bib | https://aclanthology.org/2023.emnlp-main.938/ | @inproceedings{bylinina-etal-2023-connecting,
title = "Connecting degree and polarity: An artificial language learning study",
author = "Bylinina, Lisa and
Tikhonov, Alexey and
Garmash, Ekaterina",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.938",
doi = "10.18653/v1/2023.emnlp-main.938",
pages = "15168--15177",
abstract = "We investigate a new linguistic generalisation in pre-trained language models (taking BERT Devlin et al. 2019 as a case study). We focus on degree modifiers (expressions like slightly, very, rather, extremely) and test the hypothesis that the degree expressed by a modifier (low, medium or high degree) is related to the modifier{'}s sensitivity to sentence polarity (whether it shows preference for affirmative or negative sentences or neither). To probe this connection, we apply the Artificial Language Learning experimental paradigm from psycholinguistics to a neural language model. Our experimental results suggest that BERT generalizes in line with existing linguistic observations that relate de- gree semantics to polarity sensitivity, including the main one: low degree semantics is associated with preference towards positive polarity.",
}
| We investigate a new linguistic generalisation in pre-trained language models (taking BERT Devlin et al. 2019 as a case study). We focus on degree modifiers (expressions like slightly, very, rather, extremely) and test the hypothesis that the degree expressed by a modifier (low, medium or high degree) is related to the modifier{'}s sensitivity to sentence polarity (whether it shows preference for affirmative or negative sentences or neither). To probe this connection, we apply the Artificial Language Learning experimental paradigm from psycholinguistics to a neural language model. Our experimental results suggest that BERT generalizes in line with existing linguistic observations that relate de- gree semantics to polarity sensitivity, including the main one: low degree semantics is associated with preference towards positive polarity. | [
"Bylinina, Lisa",
"Tikhonov, Alexey",
"Garmash, Ekaterina"
] | Connecting degree and polarity: An artificial language learning study | emnlp-main.938 | 2109.06333 | [
"https://github.com/altsoph/artificial_degree_modifiers"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Oral |
|
https://aclanthology.org/2023.emnlp-main.939.bib | https://aclanthology.org/2023.emnlp-main.939/ | @inproceedings{mishra-etal-2023-prompting,
title = "Prompting with Pseudo-Code Instructions",
author = "Mishra, Mayank and
Kumar, Prince and
Bhat, Riyaz and
Murthy, Rudra and
Contractor, Danish and
Tamilselvam, Srikanth",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.939",
doi = "10.18653/v1/2023.emnlp-main.939",
pages = "15178--15197",
abstract = "Prompting with natural language instructions has recently emerged as a popular method of harnessing the capabilities of large language models (LLM). Given the inherent ambiguity present in natural language, it is intuitive to consider the possible advantages of prompting with less ambiguous prompt styles, like pseudo-code. In this paper, we explore if prompting via pseudo-code instructions helps improve the performance of pre-trained language models. We manually create a dataset of pseudo-code prompts for 132 different tasks spanning classification, QA, and generative language tasks, sourced from the Super-NaturalInstructions dataset. Using these prompts along with their counterparts in natural language, we study their performance on two LLM families - BLOOM, CodeGen. Our experiments show that using pseudo-code instructions leads to better results, with an average increase (absolute) of 7-16 points in F1 scores for classification tasks and an improvement (relative) of 12-38{\%} in aggregate ROUGE-L scores across all tasks. We include detailed ablation studies which indicate that code comments, docstrings, and the structural clues encoded in pseudo-code all contribute towards the improvement in performance. To the best of our knowledge, our work is the first to demonstrate how pseudo-code prompts can be helpful in improving the performance of pre-trained LMs.",
}
| Prompting with natural language instructions has recently emerged as a popular method of harnessing the capabilities of large language models (LLM). Given the inherent ambiguity present in natural language, it is intuitive to consider the possible advantages of prompting with less ambiguous prompt styles, like pseudo-code. In this paper, we explore if prompting via pseudo-code instructions helps improve the performance of pre-trained language models. We manually create a dataset of pseudo-code prompts for 132 different tasks spanning classification, QA, and generative language tasks, sourced from the Super-NaturalInstructions dataset. Using these prompts along with their counterparts in natural language, we study their performance on two LLM families - BLOOM, CodeGen. Our experiments show that using pseudo-code instructions leads to better results, with an average increase (absolute) of 7-16 points in F1 scores for classification tasks and an improvement (relative) of 12-38{\%} in aggregate ROUGE-L scores across all tasks. We include detailed ablation studies which indicate that code comments, docstrings, and the structural clues encoded in pseudo-code all contribute towards the improvement in performance. To the best of our knowledge, our work is the first to demonstrate how pseudo-code prompts can be helpful in improving the performance of pre-trained LMs. | [
"Mishra, Mayank",
"Kumar, Prince",
"Bhat, Riyaz",
"Murthy, Rudra",
"Contractor, Danish",
"Tamilselvam, Srikanth"
] | Prompting with Pseudo-Code Instructions | emnlp-main.939 | 2305.11790 | [
"https://github.com/mayank31398/pseudo-code-instructions"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.940.bib | https://aclanthology.org/2023.emnlp-main.940/ | @inproceedings{romanou-etal-2023-crab,
title = "{CRAB}: Assessing the Strength of Causal Relationships Between Real-world Events",
author = "Romanou, Angelika and
Montariol, Syrielle and
Paul, Debjit and
Laugier, Leo and
Aberer, Karl and
Bosselut, Antoine",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.940",
doi = "10.18653/v1/2023.emnlp-main.940",
pages = "15198--15216",
abstract = "Understanding narratives requires reasoning about the cause-and-effect relationships between events mentioned in the text. While existing foundation models yield impressive results in many NLP tasks requiring reasoning, it is unclear whether they understand the complexity of the underlying network of causal relationships of events in narratives. In this work, we present CRAB, a new Causal Reasoning Assessment Benchmark designed to evaluate causal understanding of events in real-world narratives. CRAB contains fine-grained, contextual causality annotations for {\textasciitilde}2.7K pairs of real-world events that describe various newsworthy event timelines (e.g., the acquisition of Twitter by Elon Musk). Using CRAB, we measure the performance of several large language models, demonstrating that most systems achieve poor performance on the task. Motivated by classical causal principles, we also analyze the causal structures of groups of events in CRAB, and find that models perform worse on causal reasoning when events are derived from complex causal structures compared to simple linear causal chains. We make our dataset and code available to the research community.",
}
| Understanding narratives requires reasoning about the cause-and-effect relationships between events mentioned in the text. While existing foundation models yield impressive results in many NLP tasks requiring reasoning, it is unclear whether they understand the complexity of the underlying network of causal relationships of events in narratives. In this work, we present CRAB, a new Causal Reasoning Assessment Benchmark designed to evaluate causal understanding of events in real-world narratives. CRAB contains fine-grained, contextual causality annotations for {\textasciitilde}2.7K pairs of real-world events that describe various newsworthy event timelines (e.g., the acquisition of Twitter by Elon Musk). Using CRAB, we measure the performance of several large language models, demonstrating that most systems achieve poor performance on the task. Motivated by classical causal principles, we also analyze the causal structures of groups of events in CRAB, and find that models perform worse on causal reasoning when events are derived from complex causal structures compared to simple linear causal chains. We make our dataset and code available to the research community. | [
"Romanou, Angelika",
"Montariol, Syrielle",
"Paul, Debjit",
"Laugier, Leo",
"Aberer, Karl",
"Bosselut, Antoine"
] | CRAB: Assessing the Strength of Causal Relationships Between Real-world Events | emnlp-main.940 | 2311.04284 | [
"https://github.com/agromanou/crab"
] | https://huggingface.co/papers/2311.04284 | 1 | 0 | 0 | 6 | [] | [
"angelika/CRAB"
] | [] | 1 | Poster |
https://aclanthology.org/2023.emnlp-main.941.bib | https://aclanthology.org/2023.emnlp-main.941/ | @inproceedings{fung-etal-2023-normsage,
title = "{NORMSAGE}: Multi-Lingual Multi-Cultural Norm Discovery from Conversations On-the-Fly",
author = "Fung, Yi and
Chakrabarty, Tuhin and
Guo, Hao and
Rambow, Owen and
Muresan, Smaranda and
Ji, Heng",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.941",
doi = "10.18653/v1/2023.emnlp-main.941",
pages = "15217--15230",
abstract = "Knowledge of norms is needed to understand and reason about acceptable behavior in human communication and interactions across sociocultural scenarios. Most computational research on norms has focused on a single culture, and manually built datasets, from non-conversational settings. We address these limitations by proposing a new framework, NormSage, to automatically extract culture-specific norms from multi-lingual conversations. NormSage uses GPT-3 prompting to 1) extract candidate norms directly from conversations and 2) provide explainable self-verification to ensure correctness and relevance. Comprehensive empirical results show the promise of our approach to extract high-quality culture-aware norms from multi-lingual conversations (English and Chinese), across several quality metrics. Further, our relevance verification can be extended to assess the adherence and violation of any norm with respect to a conversation on-the-fly, along with textual explanation. NormSage achieves an AUC of 94.6{\%} in this grounding setup, with generated explanations matching human-written quality.",
}
| Knowledge of norms is needed to understand and reason about acceptable behavior in human communication and interactions across sociocultural scenarios. Most computational research on norms has focused on a single culture, and manually built datasets, from non-conversational settings. We address these limitations by proposing a new framework, NormSage, to automatically extract culture-specific norms from multi-lingual conversations. NormSage uses GPT-3 prompting to 1) extract candidate norms directly from conversations and 2) provide explainable self-verification to ensure correctness and relevance. Comprehensive empirical results show the promise of our approach to extract high-quality culture-aware norms from multi-lingual conversations (English and Chinese), across several quality metrics. Further, our relevance verification can be extended to assess the adherence and violation of any norm with respect to a conversation on-the-fly, along with textual explanation. NormSage achieves an AUC of 94.6{\%} in this grounding setup, with generated explanations matching human-written quality. | [
"Fung, Yi",
"Chakrabarty, Tuhin",
"Guo, Hao",
"Rambow, Owen",
"Muresan, Smar",
"a",
"Ji, Heng"
] | NORMSAGE: Multi-Lingual Multi-Cultural Norm Discovery from Conversations On-the-Fly | emnlp-main.941 | 2210.08604 | [
"https://github.com/yrf1/normsage"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.942.bib | https://aclanthology.org/2023.emnlp-main.942/ | @inproceedings{sahak-etal-2023-state,
title = "A State-Vector Framework for Dataset Effects",
author = "Sahak, Esmat and
Zhu, Zining and
Rudzicz, Frank",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.942",
doi = "10.18653/v1/2023.emnlp-main.942",
pages = "15231--15245",
abstract = "The impressive success of recent deep neural network (DNN)-based systems is significantly influenced by the high-quality datasets used in training. However, the effects of the datasets, especially how they interact with each other, remain underexplored. We propose a state-vector framework to enable rigorous studies in this direction. This framework uses idealized probing test results as the bases of a vector space. This framework allows us to quantify the effects of both standalone and interacting datasets. We show that the significant effects of some commonly-used language understanding datasets are characteristic and are concentrated on a few linguistic dimensions. Additionally, we observe some {``}spill-over{''} effects: the datasets could impact the models along dimensions that may seem unrelated to the intended tasks. Our state-vector framework paves the way for a systematic understanding of the dataset effects, a crucial component in responsible and robust model development.",
}
| The impressive success of recent deep neural network (DNN)-based systems is significantly influenced by the high-quality datasets used in training. However, the effects of the datasets, especially how they interact with each other, remain underexplored. We propose a state-vector framework to enable rigorous studies in this direction. This framework uses idealized probing test results as the bases of a vector space. This framework allows us to quantify the effects of both standalone and interacting datasets. We show that the significant effects of some commonly-used language understanding datasets are characteristic and are concentrated on a few linguistic dimensions. Additionally, we observe some {``}spill-over{''} effects: the datasets could impact the models along dimensions that may seem unrelated to the intended tasks. Our state-vector framework paves the way for a systematic understanding of the dataset effects, a crucial component in responsible and robust model development. | [
"Sahak, Esmat",
"Zhu, Zining",
"Rudzicz, Frank"
] | A State-Vector Framework for Dataset Effects | emnlp-main.942 | 2310.10955 | [
"https://github.com/esmatsahak/emnlp-2023_a-state-vector-framework-for-dataset-effects_repository"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.943.bib | https://aclanthology.org/2023.emnlp-main.943/ | @inproceedings{jin-etal-2023-challenges,
title = "Challenges in Context-Aware Neural Machine Translation",
author = "Jin, Linghao and
He, Jacqueline and
May, Jonathan and
Ma, Xuezhe",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.943",
doi = "10.18653/v1/2023.emnlp-main.943",
pages = "15246--15263",
abstract = "Context-aware neural machine translation, a paradigm that involves leveraging information beyond sentence-level context to resolve inter-sentential discourse dependencies and improve document-level translation quality, has given rise to a number of recent techniques. However, despite well-reasoned intuitions, most context-aware translation models show only modest improvements over sentence-level systems. In this work, we investigate and present several core challenges that impede progress within the field, relating to discourse phenomena, context usage, model architectures, and document-level evaluation. To address these problems, we propose a more realistic setting for document-level translation, called paragraph-to-paragraph (PARA2PARA) translation, and collect a new dataset of Chinese-English novels to promote future research.",
}
| Context-aware neural machine translation, a paradigm that involves leveraging information beyond sentence-level context to resolve inter-sentential discourse dependencies and improve document-level translation quality, has given rise to a number of recent techniques. However, despite well-reasoned intuitions, most context-aware translation models show only modest improvements over sentence-level systems. In this work, we investigate and present several core challenges that impede progress within the field, relating to discourse phenomena, context usage, model architectures, and document-level evaluation. To address these problems, we propose a more realistic setting for document-level translation, called paragraph-to-paragraph (PARA2PARA) translation, and collect a new dataset of Chinese-English novels to promote future research. | [
"Jin, Linghao",
"He, Jacqueline",
"May, Jonathan",
"Ma, Xuezhe"
] | Challenges in Context-Aware Neural Machine Translation | emnlp-main.943 | 2305.13751 | [
"https://github.com/linghao-jin/canmt-challenges"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.944.bib | https://aclanthology.org/2023.emnlp-main.944/ | @inproceedings{liu-etal-2023-task,
title = "Task-Adaptive Tokenization: Enhancing Long-Form Text Generation Efficacy in Mental Health and Beyond",
author = "Liu, Siyang and
Deng, Naihao and
Sabour, Sahand and
Jia, Yilin and
Huang, Minlie and
Mihalcea, Rada",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.944",
doi = "10.18653/v1/2023.emnlp-main.944",
pages = "15264--15281",
abstract = "We propose task-adaptive tokenization as a way to adapt the generation pipeline to the specifics of a downstream task and enhance long-form generation in mental health. Inspired by insights from cognitive science, our task-adaptive tokenizer samples variable segmentations from multiple outcomes, with sampling probabilities optimized based on task-specific data. We introduce a strategy for building a specialized vocabulary and introduce a vocabulary merging protocol that allows for the integration of task-specific tokens into the pre-trained model{'}s tokenization step. Through extensive experiments on psychological question-answering tasks in both Chinese and English, we find that our task-adaptive tokenization approach brings a significant improvement in generation performance while using up to 60{\%} fewer tokens. Preliminary experiments point to promising results when using our tokenization approach with very large language models.",
}
| We propose task-adaptive tokenization as a way to adapt the generation pipeline to the specifics of a downstream task and enhance long-form generation in mental health. Inspired by insights from cognitive science, our task-adaptive tokenizer samples variable segmentations from multiple outcomes, with sampling probabilities optimized based on task-specific data. We introduce a strategy for building a specialized vocabulary and introduce a vocabulary merging protocol that allows for the integration of task-specific tokens into the pre-trained model{'}s tokenization step. Through extensive experiments on psychological question-answering tasks in both Chinese and English, we find that our task-adaptive tokenization approach brings a significant improvement in generation performance while using up to 60{\%} fewer tokens. Preliminary experiments point to promising results when using our tokenization approach with very large language models. | [
"Liu, Siyang",
"Deng, Naihao",
"Sabour, Sah",
"",
"Jia, Yilin",
"Huang, Minlie",
"Mihalcea, Rada"
] | Task-Adaptive Tokenization: Enhancing Long-Form Text Generation Efficacy in Mental Health and Beyond | emnlp-main.944 | 2310.05317 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.945.bib | https://aclanthology.org/2023.emnlp-main.945/ | @inproceedings{chakraborty-etal-2023-factify3m,
title = "{FACTIFY}3{M}: A benchmark for multimodal fact verification with explainability through 5{W} Question-Answering",
author = "Chakraborty, Megha and
Pahwa, Khushbu and
Rani, Anku and
Chatterjee, Shreyas and
Dalal, Dwip and
Dave, Harshit and
G, Ritvik and
Gurumurthy, Preethi and
Mahor, Adarsh and
Mukherjee, Samahriti and
Pakala, Aditya and
Paul, Ishan and
Reddy, Janvita and
Sarkar, Arghya and
Sensharma, Kinjal and
Chadha, Aman and
Sheth, Amit and
Das, Amitava",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.945",
doi = "10.18653/v1/2023.emnlp-main.945",
pages = "15282--15322",
abstract = "Combating disinformation is one of the burning societal crises - about 67{\%} of the American population believes that disinformation produces a lot of uncertainty, and 10{\%} of them knowingly propagate disinformation. Evidence shows that disinformation can manipulate democratic processes and public opinion, causing disruption in the share market, panic and anxiety in society, and even death during crises. Therefore, disinformation should be identified promptly and, if possible, mitigated. With approximately 3.2 billion images and 720,000 hours of video shared online daily on social media platforms, scalable detection of multimodal disinformation requires efficient fact verification. Despite progress in automatic text-based fact verification (e.g., FEVER, LIAR), the research community lacks substantial effort in multimodal fact verification. To address this gap, we introduce FACTIFY 3M, a dataset of 3 million samples that pushes the boundaries of the domain of fact verification via a multimodal fake news dataset, in addition to offering explainability through the concept of 5W question-answering. Salient features of the dataset include: (i) textual claims, (ii) ChatGPT-generated paraphrased claims, (iii) associated images, (iv) stable diffusion-generated additional images (i.e., visual paraphrases), (v) pixel-level image heatmap to foster image-text explainability of the claim, (vi) 5W QA pairs, and (vii) adversarial fake news stories.",
}
| Combating disinformation is one of the burning societal crises - about 67{\%} of the American population believes that disinformation produces a lot of uncertainty, and 10{\%} of them knowingly propagate disinformation. Evidence shows that disinformation can manipulate democratic processes and public opinion, causing disruption in the share market, panic and anxiety in society, and even death during crises. Therefore, disinformation should be identified promptly and, if possible, mitigated. With approximately 3.2 billion images and 720,000 hours of video shared online daily on social media platforms, scalable detection of multimodal disinformation requires efficient fact verification. Despite progress in automatic text-based fact verification (e.g., FEVER, LIAR), the research community lacks substantial effort in multimodal fact verification. To address this gap, we introduce FACTIFY 3M, a dataset of 3 million samples that pushes the boundaries of the domain of fact verification via a multimodal fake news dataset, in addition to offering explainability through the concept of 5W question-answering. Salient features of the dataset include: (i) textual claims, (ii) ChatGPT-generated paraphrased claims, (iii) associated images, (iv) stable diffusion-generated additional images (i.e., visual paraphrases), (v) pixel-level image heatmap to foster image-text explainability of the claim, (vi) 5W QA pairs, and (vii) adversarial fake news stories. | [
"Chakraborty, Megha",
"Pahwa, Khushbu",
"Rani, Anku",
"Chatterjee, Shreyas",
"Dalal, Dwip",
"Dave, Harshit",
"G, Ritvik",
"Gurumurthy, Preethi",
"Mahor, Adarsh",
"Mukherjee, Samahriti",
"Pakala, Aditya",
"Paul, Ishan",
"Reddy, Janvita",
"Sarkar, Arghya",
"Sensharma, Kinjal",
"Chadha, Aman",
"Sheth, Amit",
"Das, Amitava"
] | FACTIFY3M: A benchmark for multimodal fact verification with explainability through 5W Question-Answering | emnlp-main.945 | 2306.05523 | [
""
] | https://huggingface.co/papers/2306.05523 | 1 | 0 | 0 | 18 | [] | [] | [] | 1 | Poster |
https://aclanthology.org/2023.emnlp-main.946.bib | https://aclanthology.org/2023.emnlp-main.946/ | @inproceedings{zhu-etal-2023-building,
title = "Building Multi-domain Dialog State Trackers from Single-domain Dialogs",
author = "Zhu, Qi and
Zhang, Zheng and
Zhu, Xiaoyan and
Huang, Minlie",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.946",
doi = "10.18653/v1/2023.emnlp-main.946",
pages = "15323--15335",
abstract = "Existing multi-domain dialog state tracking (DST) models are developed based on multi-domain dialogs, which require significant manual effort to define domain relations and collect data. This process can be challenging and expensive, particularly when numerous domains are involved. In this paper, we propose a divide-and-conquer (DAC) DST paradigm and a multi-domain dialog synthesis framework, which makes building multi-domain DST models from single-domain dialogs possible. The DAC paradigm segments a multi-domain dialog into multiple single-domain dialogs for DST, which makes models generalize better on dialogs involving unseen domain combinations. The multi-domain dialog synthesis framework merges several potentially related single-domain dialogs into one multi-domain dialog and modifies the dialog to simulate domain relations. The synthesized dialogs can help DST models capture the value transfer between domains. Experiments with three representative DST models on two datasets demonstrate the effectiveness of our proposed DAC paradigm and data synthesis framework.",
}
| Existing multi-domain dialog state tracking (DST) models are developed based on multi-domain dialogs, which require significant manual effort to define domain relations and collect data. This process can be challenging and expensive, particularly when numerous domains are involved. In this paper, we propose a divide-and-conquer (DAC) DST paradigm and a multi-domain dialog synthesis framework, which makes building multi-domain DST models from single-domain dialogs possible. The DAC paradigm segments a multi-domain dialog into multiple single-domain dialogs for DST, which makes models generalize better on dialogs involving unseen domain combinations. The multi-domain dialog synthesis framework merges several potentially related single-domain dialogs into one multi-domain dialog and modifies the dialog to simulate domain relations. The synthesized dialogs can help DST models capture the value transfer between domains. Experiments with three representative DST models on two datasets demonstrate the effectiveness of our proposed DAC paradigm and data synthesis framework. | [
"Zhu, Qi",
"Zhang, Zheng",
"Zhu, Xiaoyan",
"Huang, Minlie"
] | Building Multi-domain Dialog State Trackers from Single-domain Dialogs | emnlp-main.946 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
||
https://aclanthology.org/2023.emnlp-main.947.bib | https://aclanthology.org/2023.emnlp-main.947/ | @inproceedings{shi-etal-2023-specialist,
title = "Specialist or Generalist? Instruction Tuning for Specific {NLP} Tasks",
author = "Shi, Chufan and
Su, Yixuan and
Yang, Cheng and
Yang, Yujiu and
Cai, Deng",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.947",
doi = "10.18653/v1/2023.emnlp-main.947",
pages = "15336--15348",
abstract = "The potential of large language models (LLMs) to simultaneously perform a wide range of natural language processing (NLP) tasks has been the subject of extensive research. Although instruction tuning has proven to be a data-efficient method for transforming LLMs into such generalist models, their performance still lags behind specialist models trained exclusively for specific tasks. In this paper, we investigate whether incorporating broadcoverage generalist instruction tuning can contribute to building a specialist model. We hypothesize that its efficacy depends on task specificity and skill requirements. Our experiments assess four target tasks with distinct coverage levels, revealing that integrating generalist instruction tuning consistently enhances model performance when the task coverage is broad. The effect is particularly pronounced when the amount of task-specific training data is limited. Further investigation into three target tasks focusing on different capabilities demonstrates that generalist instruction tuning improves understanding and reasoning abilities. However, for tasks requiring factual knowledge, generalist data containing hallucinatory information may negatively affect the model{'}s performance. Overall, our work provides a systematic guide for developing specialist models with general instruction tuning.",
}
| The potential of large language models (LLMs) to simultaneously perform a wide range of natural language processing (NLP) tasks has been the subject of extensive research. Although instruction tuning has proven to be a data-efficient method for transforming LLMs into such generalist models, their performance still lags behind specialist models trained exclusively for specific tasks. In this paper, we investigate whether incorporating broadcoverage generalist instruction tuning can contribute to building a specialist model. We hypothesize that its efficacy depends on task specificity and skill requirements. Our experiments assess four target tasks with distinct coverage levels, revealing that integrating generalist instruction tuning consistently enhances model performance when the task coverage is broad. The effect is particularly pronounced when the amount of task-specific training data is limited. Further investigation into three target tasks focusing on different capabilities demonstrates that generalist instruction tuning improves understanding and reasoning abilities. However, for tasks requiring factual knowledge, generalist data containing hallucinatory information may negatively affect the model{'}s performance. Overall, our work provides a systematic guide for developing specialist models with general instruction tuning. | [
"Shi, Chufan",
"Su, Yixuan",
"Yang, Cheng",
"Yang, Yujiu",
"Cai, Deng"
] | Specialist or Generalist? Instruction Tuning for Specific NLP Tasks | emnlp-main.947 | 2310.15326 | [
""
] | https://huggingface.co/papers/2310.15326 | 1 | 0 | 0 | 5 | [] | [] | [] | 1 | Poster |
https://aclanthology.org/2023.emnlp-main.948.bib | https://aclanthology.org/2023.emnlp-main.948/ | @inproceedings{lee-etal-2023-making,
title = "Making Large Language Models Better Data Creators",
author = "Lee, Dong-Ho and
Pujara, Jay and
Sewak, Mohit and
White, Ryen and
Jauhar, Sujay",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.948",
doi = "10.18653/v1/2023.emnlp-main.948",
pages = "15349--15360",
abstract = "Although large language models (LLMs) have advanced the state-of-the-art in NLP significantly, deploying them for downstream applications is still challenging due to cost, responsiveness, control, or concerns around privacy and security. As such, trainable models are still the preferred option in some cases. However, these models still require human-labeled data for optimal performance, which is expensive and time-consuming to obtain. In order to address this issue, several techniques to reduce human effort involve labeling or generating data using LLMs. Although these methods are effective for certain applications, in practice they encounter difficulties in real-world scenarios. Labeling data requires careful data selection, while generating data necessitates task-specific prompt engineering. In this paper, we propose a unified data creation pipeline that requires only a single formatting example, and which is applicable to a broad range of tasks, including traditionally problematic ones with semantically devoid label spaces. In our experiments we demonstrate that instruction-following LLMs are highly cost-effective data creators, and that models trained with these data exhibit performance better than those trained with human-labeled data (by up to 17.5{\%}) on out-of-distribution evaluation, while maintaining comparable performance on in-distribution tasks. These results have important implications for the robustness of NLP systems deployed in the real-world.",
}
| Although large language models (LLMs) have advanced the state-of-the-art in NLP significantly, deploying them for downstream applications is still challenging due to cost, responsiveness, control, or concerns around privacy and security. As such, trainable models are still the preferred option in some cases. However, these models still require human-labeled data for optimal performance, which is expensive and time-consuming to obtain. In order to address this issue, several techniques to reduce human effort involve labeling or generating data using LLMs. Although these methods are effective for certain applications, in practice they encounter difficulties in real-world scenarios. Labeling data requires careful data selection, while generating data necessitates task-specific prompt engineering. In this paper, we propose a unified data creation pipeline that requires only a single formatting example, and which is applicable to a broad range of tasks, including traditionally problematic ones with semantically devoid label spaces. In our experiments we demonstrate that instruction-following LLMs are highly cost-effective data creators, and that models trained with these data exhibit performance better than those trained with human-labeled data (by up to 17.5{\%}) on out-of-distribution evaluation, while maintaining comparable performance on in-distribution tasks. These results have important implications for the robustness of NLP systems deployed in the real-world. | [
"Lee, Dong-Ho",
"Pujara, Jay",
"Sewak, Mohit",
"White, Ryen",
"Jauhar, Sujay"
] | Making Large Language Models Better Data Creators | emnlp-main.948 | 2310.20111 | [
"https://github.com/microsoft/llm-data-creation"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Oral |
|
https://aclanthology.org/2023.emnlp-main.949.bib | https://aclanthology.org/2023.emnlp-main.949/ | @inproceedings{wang-etal-2023-hallucination,
title = "Hallucination Detection for Generative Large Language Models by {B}ayesian Sequential Estimation",
author = "Wang, Xiaohua and
Yan, Yuliang and
Huang, Longtao and
Zheng, Xiaoqing and
Huang, Xuanjing",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.949",
doi = "10.18653/v1/2023.emnlp-main.949",
pages = "15361--15371",
abstract = "Large Language Models (LLMs) have made remarkable advancements in the field of natural language generation. However, the propensity of LLMs to generate inaccurate or non-factual content, termed {``}hallucinations{''}, remains a significant challenge. Current hallucination detection methods often necessitate the retrieval of great numbers of relevant evidence, thereby increasing response times. We introduce a unique framework that leverages statistical decision theory and Bayesian sequential analysis to optimize the trade-off between costs and benefits during the hallucination detection process. This approach does not require a predetermined number of observations. Instead, the analysis proceeds in a sequential manner, enabling an expeditious decision towards {``}belief{''} or {``}disbelief{''} through a stop-or-continue strategy. Extensive experiments reveal that this novel framework surpasses existing methods in both efficiency and precision of hallucination detection. Furthermore, it requires fewer retrieval steps on average, thus decreasing response times.",
}
| Large Language Models (LLMs) have made remarkable advancements in the field of natural language generation. However, the propensity of LLMs to generate inaccurate or non-factual content, termed {``}hallucinations{''}, remains a significant challenge. Current hallucination detection methods often necessitate the retrieval of great numbers of relevant evidence, thereby increasing response times. We introduce a unique framework that leverages statistical decision theory and Bayesian sequential analysis to optimize the trade-off between costs and benefits during the hallucination detection process. This approach does not require a predetermined number of observations. Instead, the analysis proceeds in a sequential manner, enabling an expeditious decision towards {``}belief{''} or {``}disbelief{''} through a stop-or-continue strategy. Extensive experiments reveal that this novel framework surpasses existing methods in both efficiency and precision of hallucination detection. Furthermore, it requires fewer retrieval steps on average, thus decreasing response times. | [
"Wang, Xiaohua",
"Yan, Yuliang",
"Huang, Longtao",
"Zheng, Xiaoqing",
"Huang, Xuanjing"
] | Hallucination Detection for Generative Large Language Models by Bayesian Sequential Estimation | emnlp-main.949 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
||
https://aclanthology.org/2023.emnlp-main.950.bib | https://aclanthology.org/2023.emnlp-main.950/ | @inproceedings{pang-etal-2023-guideline,
title = "Guideline Learning for In-Context Information Extraction",
author = "Pang, Chaoxu and
Cao, Yixuan and
Ding, Qiang and
Luo, Ping",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.950",
doi = "10.18653/v1/2023.emnlp-main.950",
pages = "15372--15389",
abstract = "Large language models (LLMs) can perform a new task by merely conditioning on task instructions and a few input-output examples, without optimizing any parameters. This is called In-Context Learning (ICL). In-context Information Extraction (IE) has recently garnered attention in the research community. However, the performance of In-context IE generally lags behind the state-of-the-art supervised expert models. We highlight a key reason for this shortfall: underspecified task description. The limited-length context struggles to thoroughly express the intricate IE task instructions and various edge cases, leading to misalignment in task comprehension with humans. In this paper, we propose a Guideline Learning (GL) framework for In-context IE which reflectively learns and follows guidelines. During the learning phrase, GL automatically synthesizes a set of guidelines based on a few error cases, and during inference, GL retrieves helpful guidelines for better ICL. Moreover, we propose a self-consistency-based active learning method to enhance the efficiency of GL. Experiments on event extraction and relation extraction show that GL can significantly improve the performance of in-context IE.",
}
| Large language models (LLMs) can perform a new task by merely conditioning on task instructions and a few input-output examples, without optimizing any parameters. This is called In-Context Learning (ICL). In-context Information Extraction (IE) has recently garnered attention in the research community. However, the performance of In-context IE generally lags behind the state-of-the-art supervised expert models. We highlight a key reason for this shortfall: underspecified task description. The limited-length context struggles to thoroughly express the intricate IE task instructions and various edge cases, leading to misalignment in task comprehension with humans. In this paper, we propose a Guideline Learning (GL) framework for In-context IE which reflectively learns and follows guidelines. During the learning phrase, GL automatically synthesizes a set of guidelines based on a few error cases, and during inference, GL retrieves helpful guidelines for better ICL. Moreover, we propose a self-consistency-based active learning method to enhance the efficiency of GL. Experiments on event extraction and relation extraction show that GL can significantly improve the performance of in-context IE. | [
"Pang, Chaoxu",
"Cao, Yixuan",
"Ding, Qiang",
"Luo, Ping"
] | Guideline Learning for In-Context Information Extraction | emnlp-main.950 | 2310.05066 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Oral |
|
https://aclanthology.org/2023.emnlp-main.951.bib | https://aclanthology.org/2023.emnlp-main.951/ | @inproceedings{dong-etal-2023-open,
title = "Open Information Extraction via Chunks",
author = "Dong, Kuicai and
Sun, Aixin and
Kim, Jung-jae and
Li, Xiaoli",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.951",
doi = "10.18653/v1/2023.emnlp-main.951",
pages = "15390--15404",
abstract = "Open Information Extraction (OIE) aims to extract relational tuples from open-domain sentences. Existing OIE systems split a sentence into tokens and recognize token spans as tuple relations and arguments. We instead propose Sentence as Chunk sequence (SaC) and recognize chunk spans as tuple relations and arguments. We argue that SaC has better properties for OIE than sentence as token sequence, and evaluate four choices of chunks (i.e., CoNLL chunks, OIA simple phrases, noun phrases, and spans from SpanOIE). Also, we propose a simple end-to-end BERT-based model, Chunk-OIE, for sentence chunking and tuple extraction on top of SaC. Chunk-OIE achieves state-of-the-art results on multiple OIE datasets, showing that SaC benefits the OIE task.",
}
| Open Information Extraction (OIE) aims to extract relational tuples from open-domain sentences. Existing OIE systems split a sentence into tokens and recognize token spans as tuple relations and arguments. We instead propose Sentence as Chunk sequence (SaC) and recognize chunk spans as tuple relations and arguments. We argue that SaC has better properties for OIE than sentence as token sequence, and evaluate four choices of chunks (i.e., CoNLL chunks, OIA simple phrases, noun phrases, and spans from SpanOIE). Also, we propose a simple end-to-end BERT-based model, Chunk-OIE, for sentence chunking and tuple extraction on top of SaC. Chunk-OIE achieves state-of-the-art results on multiple OIE datasets, showing that SaC benefits the OIE task. | [
"Dong, Kuicai",
"Sun, Aixin",
"Kim, Jung-jae",
"Li, Xiaoli"
] | Open Information Extraction via Chunks | emnlp-main.951 | 2305.03299 | [
"https://github.com/daviddongkc/chunk_oie"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.952.bib | https://aclanthology.org/2023.emnlp-main.952/ | @inproceedings{chen-etal-2023-rethinking,
title = "Rethinking Word-Level Auto-Completion in Computer-Aided Translation",
author = "Chen, Xingyu and
Liu, Lemao and
Huang, Guoping and
Zhang, Zhirui and
Yang, Mingming and
Shi, Shuming and
Wang, Rui",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.952",
doi = "10.18653/v1/2023.emnlp-main.952",
pages = "15405--15415",
abstract = "Word-level auto-completion (WLAC) plays a crucial role in Computer-Assisted Translation. While previous studies have primarily focused on designing complex model architectures, this paper takes a different perspective by rethinking the fundamental question: what kind of words are good auto-completions? We introduce a measurable criterion to address this question and discover that existing WLAC models often fail to meet this criterion. Building upon this observation, we propose an effective approach to enhance WLAC performance by promoting adherence to the criterion. Notably, the proposed approach is general and can be applied to various encoder-based architectures. Through extensive experiments, we demonstrate that our approach outperforms the top-performing system submitted to the WLAC shared tasks in WMT2022, while utilizing significantly smaller model sizes.",
}
| Word-level auto-completion (WLAC) plays a crucial role in Computer-Assisted Translation. While previous studies have primarily focused on designing complex model architectures, this paper takes a different perspective by rethinking the fundamental question: what kind of words are good auto-completions? We introduce a measurable criterion to address this question and discover that existing WLAC models often fail to meet this criterion. Building upon this observation, we propose an effective approach to enhance WLAC performance by promoting adherence to the criterion. Notably, the proposed approach is general and can be applied to various encoder-based architectures. Through extensive experiments, we demonstrate that our approach outperforms the top-performing system submitted to the WLAC shared tasks in WMT2022, while utilizing significantly smaller model sizes. | [
"Chen, Xingyu",
"Liu, Lemao",
"Huang, Guoping",
"Zhang, Zhirui",
"Yang, Mingming",
"Shi, Shuming",
"Wang, Rui"
] | Rethinking Word-Level Auto-Completion in Computer-Aided Translation | emnlp-main.952 | 2310.14523 | [
"https://github.com/galaxychen/wlac-joint-training"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.953.bib | https://aclanthology.org/2023.emnlp-main.953/ | @inproceedings{arias-etal-2023-automatic,
title = "Automatic Transcription of Handwritten Old {O}ccitan Language",
author = {Garces Arias, Esteban and
Pai, Vallari and
Sch{\"o}ffel, Matthias and
Heumann, Christian and
A{\ss}enmacher, Matthias},
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.953",
doi = "10.18653/v1/2023.emnlp-main.953",
pages = "15416--15439",
abstract = "While existing neural network-based approaches have shown promising results in Handwritten Text Recognition (HTR) for high-resource languages and standardized/machine-written text, their application to low-resource languages often presents challenges, resulting in reduced effectiveness. In this paper, we propose an innovative HTR approach that leverages the Transformer architecture for recognizing handwritten Old Occitan language. Given the limited availability of data, which comprises only word pairs of graphical variants and lemmas, we develop and rely on elaborate data augmentation techniques for both text and image data. Our model combines a custom-trained Swin image encoder with a BERT text decoder, which we pre-train using a large-scale augmented synthetic data set and fine-tune on the small human-labeled data set. Experimental results reveal that our approach surpasses the performance of current state-of-the-art models for Old Occitan HTR, including open-source Transformer-based models such as a fine-tuned TrOCR and commercial applications like Google Cloud Vision. To nurture further research and development, we make our models, data sets, and code publicly available.",
}
| While existing neural network-based approaches have shown promising results in Handwritten Text Recognition (HTR) for high-resource languages and standardized/machine-written text, their application to low-resource languages often presents challenges, resulting in reduced effectiveness. In this paper, we propose an innovative HTR approach that leverages the Transformer architecture for recognizing handwritten Old Occitan language. Given the limited availability of data, which comprises only word pairs of graphical variants and lemmas, we develop and rely on elaborate data augmentation techniques for both text and image data. Our model combines a custom-trained Swin image encoder with a BERT text decoder, which we pre-train using a large-scale augmented synthetic data set and fine-tune on the small human-labeled data set. Experimental results reveal that our approach surpasses the performance of current state-of-the-art models for Old Occitan HTR, including open-source Transformer-based models such as a fine-tuned TrOCR and commercial applications like Google Cloud Vision. To nurture further research and development, we make our models, data sets, and code publicly available. | [
"Garces Arias, Esteban",
"Pai, Vallari",
"Sch{\\\"o}ffel, Matthias",
"Heumann, Christian",
"A{\\ss}enmacher, Matthias"
] | Automatic Transcription of Handwritten Old Occitan Language | emnlp-main.953 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Oral |
||
https://aclanthology.org/2023.emnlp-main.954.bib | https://aclanthology.org/2023.emnlp-main.954/ | @inproceedings{xu-etal-2023-corefprompt,
title = "{C}oref{P}rompt: Prompt-based Event Coreference Resolution by Measuring Event Type and Argument Compatibilities",
author = "Xu, Sheng and
Li, Peifeng and
Zhu, Qiaoming",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.954",
doi = "10.18653/v1/2023.emnlp-main.954",
pages = "15440--15452",
abstract = "Event coreference resolution (ECR) aims to group event mentions referring to the same real-world event into clusters. Most previous studies adopt the {``}encoding first, then scoring{''} framework, making the coreference judgment rely on event encoding. Furthermore, current methods struggle to leverage human-summarized ECR rules, e.g., coreferential events should have the same event type, to guide the model. To address these two issues, we propose a prompt-based approach, CorefPrompt, to transform ECR into a cloze-style MLM (masked language model) task. This allows for simultaneous event modeling and coreference discrimination within a single template, with a fully shared context. In addition, we introduce two auxiliary prompt tasks, event-type compatibility and argument compatibility, to explicitly demonstrate the reasoning process of ECR, which helps the model make final predictions. Experimental results show that our method CorefPrompt performs well in a state-of-the-art (SOTA) benchmark.",
}
| Event coreference resolution (ECR) aims to group event mentions referring to the same real-world event into clusters. Most previous studies adopt the {``}encoding first, then scoring{''} framework, making the coreference judgment rely on event encoding. Furthermore, current methods struggle to leverage human-summarized ECR rules, e.g., coreferential events should have the same event type, to guide the model. To address these two issues, we propose a prompt-based approach, CorefPrompt, to transform ECR into a cloze-style MLM (masked language model) task. This allows for simultaneous event modeling and coreference discrimination within a single template, with a fully shared context. In addition, we introduce two auxiliary prompt tasks, event-type compatibility and argument compatibility, to explicitly demonstrate the reasoning process of ECR, which helps the model make final predictions. Experimental results show that our method CorefPrompt performs well in a state-of-the-art (SOTA) benchmark. | [
"Xu, Sheng",
"Li, Peifeng",
"Zhu, Qiaoming"
] | CorefPrompt: Prompt-based Event Coreference Resolution by Measuring Event Type and Argument Compatibilities | emnlp-main.954 | 2310.14512 | [
"https://github.com/jsksxs360/prompt-event-coref-emnlp2023"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Oral |
|
https://aclanthology.org/2023.emnlp-main.955.bib | https://aclanthology.org/2023.emnlp-main.955/ | @inproceedings{lu-etal-2023-anaphor,
title = "Anaphor Assisted Document-Level Relation Extraction",
author = "Lu, Chonggang and
Zhang, Richong and
Sun, Kai and
Kim, Jaein and
Zhang, Cunwang and
Mao, Yongyi",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.955",
doi = "10.18653/v1/2023.emnlp-main.955",
pages = "15453--15464",
abstract = "Document-level relation extraction (DocRE) involves identifying relations between entities distributed in multiple sentences within a document. Existing methods focus on building a heterogeneous document graph to model the internal structure of an entity and the external interaction between entities. However, there are two drawbacks in existing methods. On one hand, anaphor plays an important role in reasoning to identify relations between entities but is ignored by these methods. On the other hand, these methods achieve cross-sentence entity interactions implicitly by utilizing a document or sentences as intermediate nodes. Such an approach has difficulties in learning fine-grained interactions between entities across different sentences, resulting in sub-optimal performance. To address these issues, we propose an Anaphor-Assisted (AA) framework for DocRE tasks. Experimental results on the widely-used datasets demonstrate that our model achieves a new state-of-the-art performance.",
}
| Document-level relation extraction (DocRE) involves identifying relations between entities distributed in multiple sentences within a document. Existing methods focus on building a heterogeneous document graph to model the internal structure of an entity and the external interaction between entities. However, there are two drawbacks in existing methods. On one hand, anaphor plays an important role in reasoning to identify relations between entities but is ignored by these methods. On the other hand, these methods achieve cross-sentence entity interactions implicitly by utilizing a document or sentences as intermediate nodes. Such an approach has difficulties in learning fine-grained interactions between entities across different sentences, resulting in sub-optimal performance. To address these issues, we propose an Anaphor-Assisted (AA) framework for DocRE tasks. Experimental results on the widely-used datasets demonstrate that our model achieves a new state-of-the-art performance. | [
"Lu, Chonggang",
"Zhang, Richong",
"Sun, Kai",
"Kim, Jaein",
"Zhang, Cunwang",
"Mao, Yongyi"
] | Anaphor Assisted Document-Level Relation Extraction | emnlp-main.955 | 2310.18604 | [
"https://github.com/burgerburgerburger/aa"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Oral |
|
https://aclanthology.org/2023.emnlp-main.956.bib | https://aclanthology.org/2023.emnlp-main.956/ | @inproceedings{tang-etal-2023-finentity,
title = "{F}in{E}ntity: Entity-level Sentiment Classification for Financial Texts",
author = "Tang, Yixuan and
Yang, Yi and
Huang, Allen and
Tam, Andy and
Tang, Justin",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.956",
doi = "10.18653/v1/2023.emnlp-main.956",
pages = "15465--15471",
abstract = "In the financial domain, conducting entity-level sentiment analysis is crucial for accurately assessing the sentiment directed toward a specific financial entity. To our knowledge, no publicly available dataset currently exists for this purpose. In this work, we introduce an entity-level sentiment classification dataset, called FinEntity, that annotates financial entity spans and their sentiment (positive, neutral, and negative) in financial news. We document the dataset construction process in the paper. Additionally, we benchmark several pre-trained models (BERT, FinBERT, etc.) and ChatGPT on entity-level sentiment classification. In a case study, we demonstrate the practical utility of using FinEntity in monitoring cryptocurrency markets. The data and code of FinEntity is available at https://github.com/yixuantt/FinEntity.",
}
| In the financial domain, conducting entity-level sentiment analysis is crucial for accurately assessing the sentiment directed toward a specific financial entity. To our knowledge, no publicly available dataset currently exists for this purpose. In this work, we introduce an entity-level sentiment classification dataset, called FinEntity, that annotates financial entity spans and their sentiment (positive, neutral, and negative) in financial news. We document the dataset construction process in the paper. Additionally, we benchmark several pre-trained models (BERT, FinBERT, etc.) and ChatGPT on entity-level sentiment classification. In a case study, we demonstrate the practical utility of using FinEntity in monitoring cryptocurrency markets. The data and code of FinEntity is available at https://github.com/yixuantt/FinEntity. | [
"Tang, Yixuan",
"Yang, Yi",
"Huang, Allen",
"Tam, Andy",
"Tang, Justin"
] | FinEntity: Entity-level Sentiment Classification for Financial Texts | emnlp-main.956 | 2310.12406 | [
"https://github.com/yixuantt/finentity"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.957.bib | https://aclanthology.org/2023.emnlp-main.957/ | @inproceedings{liu-etal-2023-things,
title = "All Things Considered: Detecting Partisan Events from News Media with Cross-Article Comparison",
author = "Liu, Yujian and
Zhang, Xinliang and
Zou, Kaijian and
Huang, Ruihong and
Beauchamp, Nicholas and
Wang, Lu",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.957",
doi = "10.18653/v1/2023.emnlp-main.957",
pages = "15472--15488",
abstract = "Public opinion is shaped by the information news media provide, and that information in turn may be shaped by the ideological preferences of media outlets. But while much attention has been devoted to media bias via overt ideological language or topic selection, a more unobtrusive way in which the media shape opinion is via the strategic inclusion or omission of $\textit{partisan events}$ that may $\textit{support}$ one side or the other. We develop a latent variable-based framework to predict the ideology of news articles by comparing multiple articles on the same story and identifying partisan events whose inclusion or omission reveals ideology. Our experiments first validate the existence of partisan event selection, and then show that article alignment and cross-document comparison detect partisan events and article ideology better than competitive baselines. Our results reveal the high-level form of media bias, which is present even among mainstream media with strong norms of objectivity and nonpartisanship. Our codebase and dataset are available at https://github.com/launchnlp/ATC.",
}
| Public opinion is shaped by the information news media provide, and that information in turn may be shaped by the ideological preferences of media outlets. But while much attention has been devoted to media bias via overt ideological language or topic selection, a more unobtrusive way in which the media shape opinion is via the strategic inclusion or omission of $\textit{partisan events}$ that may $\textit{support}$ one side or the other. We develop a latent variable-based framework to predict the ideology of news articles by comparing multiple articles on the same story and identifying partisan events whose inclusion or omission reveals ideology. Our experiments first validate the existence of partisan event selection, and then show that article alignment and cross-document comparison detect partisan events and article ideology better than competitive baselines. Our results reveal the high-level form of media bias, which is present even among mainstream media with strong norms of objectivity and nonpartisanship. Our codebase and dataset are available at https://github.com/launchnlp/ATC. | [
"Liu, Yujian",
"Zhang, Xinliang",
"Zou, Kaijian",
"Huang, Ruihong",
"Beauchamp, Nicholas",
"Wang, Lu"
] | All Things Considered: Detecting Partisan Events from News Media with Cross-Article Comparison | emnlp-main.957 | 2310.18827 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.958.bib | https://aclanthology.org/2023.emnlp-main.958/ | @inproceedings{xiong-etal-2023-rationale,
title = "Rationale-Enhanced Language Models are Better Continual Relation Learners",
author = "Xiong, Weimin and
Song, Yifan and
Wang, Peiyi and
Li, Sujian",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.958",
doi = "10.18653/v1/2023.emnlp-main.958",
pages = "15489--15497",
abstract = "Continual relation extraction (CRE) aims to solve the problem of catastrophic forgetting when learning a sequence of newly emerging relations. Recent CRE studies have found that catastrophic forgetting arises from the model{'}s lack of robustness against future analogous relations. To address the issue, we introduce rationale, i.e., the explanations of relation classification results generated by Large Language Models (LLM), into CRE task. Specifically, we design the multi-task rationale tuning strategy to help the model learn current relations robustly. We also conduct contrastive rationale replay to further distinguish analogous relations. Experimental results on two standard benchmarks demonstrate that our method outperforms the state-of-the-art CRE models.",
}
| Continual relation extraction (CRE) aims to solve the problem of catastrophic forgetting when learning a sequence of newly emerging relations. Recent CRE studies have found that catastrophic forgetting arises from the model{'}s lack of robustness against future analogous relations. To address the issue, we introduce rationale, i.e., the explanations of relation classification results generated by Large Language Models (LLM), into CRE task. Specifically, we design the multi-task rationale tuning strategy to help the model learn current relations robustly. We also conduct contrastive rationale replay to further distinguish analogous relations. Experimental results on two standard benchmarks demonstrate that our method outperforms the state-of-the-art CRE models. | [
"Xiong, Weimin",
"Song, Yifan",
"Wang, Peiyi",
"Li, Sujian"
] | Rationale-Enhanced Language Models are Better Continual Relation Learners | emnlp-main.958 | 2310.06547 | [
"https://github.com/weiminxiong/rationalecl"
] | https://huggingface.co/papers/2310.06547 | 0 | 0 | 0 | 4 | [] | [] | [] | 1 | Poster |
https://aclanthology.org/2023.emnlp-main.959.bib | https://aclanthology.org/2023.emnlp-main.959/ | @inproceedings{das-mukherjee-2023-banglaabusememe,
title = "{B}angla{A}buse{M}eme: A Dataset for {B}engali Abusive Meme Classification",
author = "Das, Mithun and
Mukherjee, Animesh",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.959",
doi = "10.18653/v1/2023.emnlp-main.959",
pages = "15498--15512",
abstract = "The dramatic increase in the use of social media platforms for information sharing has also fueled a steep growth in online abuse. A simple yet effective way of abusing individuals or communities is by creating memes, which often integrate an image with a short piece of text layered on top of it. Such harmful elements are in rampant use and are a threat to online safety. Hence it is necessary to develop efficient models to detect and flag abusive memes. The problem becomes more challenging in a low-resource setting (e.g., Bengali memes, i.e., images with Bengali text embedded on it) because of the absence of benchmark datasets on which AI models could be trained. In this paper we bridge this gap by building a Bengali meme dataset. To setup an effective benchmark we implement several baseline models for classifying abusive memes using this dataset. We observe that multimodal models that use both textual and visual information outperform unimodal models. Our best-performing model achieves a macro F1 score of 70.51. Finally, we perform a qualitative error analysis of the misclassified memes of the best-performing text-based, image-based and multimodal models.",
}
| The dramatic increase in the use of social media platforms for information sharing has also fueled a steep growth in online abuse. A simple yet effective way of abusing individuals or communities is by creating memes, which often integrate an image with a short piece of text layered on top of it. Such harmful elements are in rampant use and are a threat to online safety. Hence it is necessary to develop efficient models to detect and flag abusive memes. The problem becomes more challenging in a low-resource setting (e.g., Bengali memes, i.e., images with Bengali text embedded on it) because of the absence of benchmark datasets on which AI models could be trained. In this paper we bridge this gap by building a Bengali meme dataset. To setup an effective benchmark we implement several baseline models for classifying abusive memes using this dataset. We observe that multimodal models that use both textual and visual information outperform unimodal models. Our best-performing model achieves a macro F1 score of 70.51. Finally, we perform a qualitative error analysis of the misclassified memes of the best-performing text-based, image-based and multimodal models. | [
"Das, Mithun",
"Mukherjee, Animesh"
] | BanglaAbuseMeme: A Dataset for Bengali Abusive Meme Classification | emnlp-main.959 | 2310.11748 | [
"https://github.com/hate-alert/banglaabusememe"
] | https://huggingface.co/papers/2310.11748 | 0 | 0 | 0 | 2 | [] | [] | [] | 1 | Oral |
https://aclanthology.org/2023.emnlp-main.960.bib | https://aclanthology.org/2023.emnlp-main.960/ | @inproceedings{bolliger-etal-2023-scandl,
title = "{S}can{DL}: A Diffusion Model for Generating Synthetic Scanpaths on Texts",
author = {Bolliger, Lena and
Reich, David and
Haller, Patrick and
Jakobi, Deborah and
Prasse, Paul and
J{\"a}ger, Lena},
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.960",
doi = "10.18653/v1/2023.emnlp-main.960",
pages = "15513--15538",
abstract = "Eye movements in reading play a crucial role in psycholinguistic research studying the cognitive mechanisms underlying human language processing. More recently, the tight coupling between eye movements and cognition has also been leveraged for language-related machine learning tasks such as the interpretability, enhancement, and pre-training of language models, as well as the inference of reader- and text-specific properties. However, scarcity of eye movement data and its unavailability at application time poses a major challenge for this line of research. Initially, this problem was tackled by resorting to cognitive models for synthesizing eye movement data. However, for the sole purpose of generating human-like scanpaths, purely data-driven machine-learning-based methods have proven to be more suitable. Following recent advances in adapting diffusion processes to discrete data, we propose ScanDL, a novel discrete sequence-to-sequence diffusion model that generates synthetic scanpaths on texts. By leveraging pre-trained word representations and jointly embedding both the stimulus text and the fixation sequence, our model captures multi-modal interactions between the two inputs. We evaluate ScanDL within- and across-dataset and demonstrate that it significantly outperforms state-of-the-art scanpath generation methods. Finally, we provide an extensive psycholinguistic analysis that underlines the model{'}s ability to exhibit human-like reading behavior. Our implementation is made available at https://github.com/DiLi-Lab/ScanDL.",
}
| Eye movements in reading play a crucial role in psycholinguistic research studying the cognitive mechanisms underlying human language processing. More recently, the tight coupling between eye movements and cognition has also been leveraged for language-related machine learning tasks such as the interpretability, enhancement, and pre-training of language models, as well as the inference of reader- and text-specific properties. However, scarcity of eye movement data and its unavailability at application time poses a major challenge for this line of research. Initially, this problem was tackled by resorting to cognitive models for synthesizing eye movement data. However, for the sole purpose of generating human-like scanpaths, purely data-driven machine-learning-based methods have proven to be more suitable. Following recent advances in adapting diffusion processes to discrete data, we propose ScanDL, a novel discrete sequence-to-sequence diffusion model that generates synthetic scanpaths on texts. By leveraging pre-trained word representations and jointly embedding both the stimulus text and the fixation sequence, our model captures multi-modal interactions between the two inputs. We evaluate ScanDL within- and across-dataset and demonstrate that it significantly outperforms state-of-the-art scanpath generation methods. Finally, we provide an extensive psycholinguistic analysis that underlines the model{'}s ability to exhibit human-like reading behavior. Our implementation is made available at https://github.com/DiLi-Lab/ScanDL. | [
"Bolliger, Lena",
"Reich, David",
"Haller, Patrick",
"Jakobi, Deborah",
"Prasse, Paul",
"J{\\\"a}ger, Lena"
] | ScanDL: A Diffusion Model for Generating Synthetic Scanpaths on Texts | emnlp-main.960 | 2310.15587 | [
"https://github.com/dili-lab/scandl"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.961.bib | https://aclanthology.org/2023.emnlp-main.961/ | @inproceedings{kang-etal-2023-values,
title = "From Values to Opinions: Predicting Human Behaviors and Stances Using Value-Injected Large Language Models",
author = "Kang, Dongjun and
Park, Joonsuk and
Jo, Yohan and
Bak, JinYeong",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.961",
doi = "10.18653/v1/2023.emnlp-main.961",
pages = "15539--15559",
abstract = "Being able to predict people{'}s opinions on issues and behaviors in realistic scenarios can be helpful in various domains, such as politics and marketing. However, conducting large-scale surveys like the European Social Survey to solicit people{'}s opinions on individual issues can incur prohibitive costs. Leveraging prior research showing influence of core human values on individual decisions and actions, we propose to use value-injected large language models (LLM) to predict opinions and behaviors. To this end, we present Value Injection Method (VIM), a collection of two methods{---}argument generation and question answering{---}designed to inject targeted value distributions into LLMs via fine-tuning. We then conduct a series of experiments on four tasks to test the effectiveness of VIM and the possibility of using value-injected LLMs to predict opinions and behaviors of people. We find that LLMs value-injected with variations of VIM substantially outperform the baselines. Also, the results suggest that opinions and behaviors can be better predicted using value-injected LLMs than the baseline approaches.",
}
| Being able to predict people{'}s opinions on issues and behaviors in realistic scenarios can be helpful in various domains, such as politics and marketing. However, conducting large-scale surveys like the European Social Survey to solicit people{'}s opinions on individual issues can incur prohibitive costs. Leveraging prior research showing influence of core human values on individual decisions and actions, we propose to use value-injected large language models (LLM) to predict opinions and behaviors. To this end, we present Value Injection Method (VIM), a collection of two methods{---}argument generation and question answering{---}designed to inject targeted value distributions into LLMs via fine-tuning. We then conduct a series of experiments on four tasks to test the effectiveness of VIM and the possibility of using value-injected LLMs to predict opinions and behaviors of people. We find that LLMs value-injected with variations of VIM substantially outperform the baselines. Also, the results suggest that opinions and behaviors can be better predicted using value-injected LLMs than the baseline approaches. | [
"Kang, Dongjun",
"Park, Joonsuk",
"Jo, Yohan",
"Bak, JinYeong"
] | From Values to Opinions: Predicting Human Behaviors and Stances Using Value-Injected Large Language Models | emnlp-main.961 | 2310.17857 | [
"https://github.com/dongjunkang/vim"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.962.bib | https://aclanthology.org/2023.emnlp-main.962/ | @inproceedings{pial-etal-2023-analyzing,
title = "Analyzing Film Adaptation through Narrative Alignment",
author = "Pial, Tanzir and
Aunti, Shahreen and
Pethe, Charuta and
Kim, Allen and
Skiena, Steven",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.962",
doi = "10.18653/v1/2023.emnlp-main.962",
pages = "15560--15579",
abstract = "Novels are often adapted into feature films, but the differences between the two media usually require dropping sections of the source text from the movie script. Here we study this screen adaptation process by constructing narrative alignments using the Smith-Waterman local alignment algorithm coupled with SBERT embedding distance to quantify text similarity between scenes and book units. We use these alignments to perform an automated analysis of 40 adaptations, revealing insights into the screenwriting process concerning (i) faithfulness of adaptation, (ii) importance of dialog, (iii) preservation of narrative order, and (iv) gender representation issues reflective of the Bechdel test.",
}
| Novels are often adapted into feature films, but the differences between the two media usually require dropping sections of the source text from the movie script. Here we study this screen adaptation process by constructing narrative alignments using the Smith-Waterman local alignment algorithm coupled with SBERT embedding distance to quantify text similarity between scenes and book units. We use these alignments to perform an automated analysis of 40 adaptations, revealing insights into the screenwriting process concerning (i) faithfulness of adaptation, (ii) importance of dialog, (iii) preservation of narrative order, and (iv) gender representation issues reflective of the Bechdel test. | [
"Pial, Tanzir",
"Aunti, Shahreen",
"Pethe, Charuta",
"Kim, Allen",
"Skiena, Steven"
] | Analyzing Film Adaptation through Narrative Alignment | emnlp-main.962 | 2311.04020 | [
"https://github.com/tanzir5/alignment_tool2.0"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.963.bib | https://aclanthology.org/2023.emnlp-main.963/ | @inproceedings{wei-etal-2023-inverse,
title = "Inverse Scaling Can Become {U}-Shaped",
author = "Wei, Jason and
Kim, Najoung and
Tay, Yi and
Le, Quoc",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.963",
doi = "10.18653/v1/2023.emnlp-main.963",
pages = "15580--15591",
abstract = "Scaling up language models has been empirically shown to improve performance on a wide range of downstream tasks. However, if we were to observe worse performance as a function of scale (inverse scaling) on certain tasks, this would indicate that scaling can also encourage behaviors that are misaligned with human preferences. The Inverse Scaling Prize (McKenzie et al. 2023) identified eleven such inverse scaling tasks, evaluated on models of up to 280B parameters and up to 500 zettaFLOPs of training compute. This paper takes a closer look at these inverse scaling tasks. In this paper, we evaluate models of up to 540B parameters, trained on five times more compute than those evaluated in the Inverse Scaling Prize. With this increased range of model sizes and compute, only four out of the eleven tasks remain inverse scaling. Six tasks exhibit U-shaped scaling, where performance decreases up to a certain size, and then increases again up to the largest model evaluated (the one remaining task displays positive scaling). In addition, 1-shot examples and chain-of-thought can help mitigate undesirable scaling patterns even further. U-shaped scaling suggests that the inverse scaling trend observed in McKenzie et al. (2023) may not continue to hold for larger models, which we attribute to the presence of distractor tasks that only sufficiently large models can avoid.",
}
| Scaling up language models has been empirically shown to improve performance on a wide range of downstream tasks. However, if we were to observe worse performance as a function of scale (inverse scaling) on certain tasks, this would indicate that scaling can also encourage behaviors that are misaligned with human preferences. The Inverse Scaling Prize (McKenzie et al. 2023) identified eleven such inverse scaling tasks, evaluated on models of up to 280B parameters and up to 500 zettaFLOPs of training compute. This paper takes a closer look at these inverse scaling tasks. In this paper, we evaluate models of up to 540B parameters, trained on five times more compute than those evaluated in the Inverse Scaling Prize. With this increased range of model sizes and compute, only four out of the eleven tasks remain inverse scaling. Six tasks exhibit U-shaped scaling, where performance decreases up to a certain size, and then increases again up to the largest model evaluated (the one remaining task displays positive scaling). In addition, 1-shot examples and chain-of-thought can help mitigate undesirable scaling patterns even further. U-shaped scaling suggests that the inverse scaling trend observed in McKenzie et al. (2023) may not continue to hold for larger models, which we attribute to the presence of distractor tasks that only sufficiently large models can avoid. | [
"Wei, Jason",
"Kim, Najoung",
"Tay, Yi",
"Le, Quoc"
] | Inverse Scaling Can Become U-Shaped | emnlp-main.963 | 2211.02011 | [
""
] | https://huggingface.co/papers/2211.02011 | 1 | 0 | 0 | 4 | [] | [] | [] | 1 | Oral |
https://aclanthology.org/2023.emnlp-main.964.bib | https://aclanthology.org/2023.emnlp-main.964/ | @inproceedings{gao-etal-2023-nearest,
title = "Nearest Neighbor Machine Translation is Meta-Optimizer on Output Projection Layer",
author = "Gao, Ruize and
Zhang, Zhirui and
Du, Yichao and
Liu, Lemao and
Wang, Rui",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.964",
doi = "10.18653/v1/2023.emnlp-main.964",
pages = "15592--15608",
abstract = "Nearest Neighbor Machine Translation ($k$NN-MT) has achieved great success in domain adaptation tasks by integrating pre-trained Neural Machine Translation (NMT) models with domain-specific token-level retrieval. However, the reasons underlying its success have not been thoroughly investigated. In this paper, we comprehensively analyze $k$NN-MT through theoretical and empirical studies. Initially, we provide new insights into the working mechanism of $k$NN-MT as an efficient technique to implicitly execute gradient descent on the output projection layer of NMT, indicating that it is a specific case of model fine-tuning. Subsequently, we conduct multi-domain experiments and word-level analysis to examine the differences in performance between $k$NN-MT and entire-model fine-tuning. Our findings suggest that: ($i$) Incorporating $k$NN-MT with adapters yields comparable translation performance to fine-tuning on in-domain test sets, while achieving better performance on out-of-domain test sets; ($ii$) Fine-tuning significantly outperforms $k$NN-MT on the recall of in-domain low-frequency words, but this gap could be bridged by optimizing the context representations with additional adapter layers.",
}
| Nearest Neighbor Machine Translation ($k$NN-MT) has achieved great success in domain adaptation tasks by integrating pre-trained Neural Machine Translation (NMT) models with domain-specific token-level retrieval. However, the reasons underlying its success have not been thoroughly investigated. In this paper, we comprehensively analyze $k$NN-MT through theoretical and empirical studies. Initially, we provide new insights into the working mechanism of $k$NN-MT as an efficient technique to implicitly execute gradient descent on the output projection layer of NMT, indicating that it is a specific case of model fine-tuning. Subsequently, we conduct multi-domain experiments and word-level analysis to examine the differences in performance between $k$NN-MT and entire-model fine-tuning. Our findings suggest that: ($i$) Incorporating $k$NN-MT with adapters yields comparable translation performance to fine-tuning on in-domain test sets, while achieving better performance on out-of-domain test sets; ($ii$) Fine-tuning significantly outperforms $k$NN-MT on the recall of in-domain low-frequency words, but this gap could be bridged by optimizing the context representations with additional adapter layers. | [
"Gao, Ruize",
"Zhang, Zhirui",
"Du, Yichao",
"Liu, Lemao",
"Wang, Rui"
] | Nearest Neighbor Machine Translation is Meta-Optimizer on Output Projection Layer | emnlp-main.964 | 2305.13034 | [
"https://github.com/ruizgao/knnmt-meta-optimizer"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.965.bib | https://aclanthology.org/2023.emnlp-main.965/ | @inproceedings{nagata-etal-2023-variance,
title = "Variance Matters: Detecting Semantic Differences without Corpus/Word Alignment",
author = "Nagata, Ryo and
Takamura, Hiroya and
Otani, Naoki and
Kawasaki, Yoshifumi",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.965",
doi = "10.18653/v1/2023.emnlp-main.965",
pages = "15609--15622",
abstract = "In this paper, we propose methods for discovering semantic differences in words appearing in two corpora. The key idea is to measure the coverage of meanings of a word in a corpus through the norm of its mean word vector, which is equivalent to examining a kind of variance of the word vector distribution. The proposed methods do not require alignments between words and/or corpora for comparison that previous methods do. All they require are to compute variance (or norms of mean word vectors) for each word type. Nevertheless, they rival the best-performing system in the SemEval-2020 Task 1. In addition, they are (i) robust for the skew in corpus sizes; (ii) capable of detecting semantic differences in infrequent words; and (iii) effective in pinpointing word instances that have a meaning missing in one of the two corpora under comparison. We show these advantages for historical corpora and also for native/non-native English corpora.",
}
| In this paper, we propose methods for discovering semantic differences in words appearing in two corpora. The key idea is to measure the coverage of meanings of a word in a corpus through the norm of its mean word vector, which is equivalent to examining a kind of variance of the word vector distribution. The proposed methods do not require alignments between words and/or corpora for comparison that previous methods do. All they require are to compute variance (or norms of mean word vectors) for each word type. Nevertheless, they rival the best-performing system in the SemEval-2020 Task 1. In addition, they are (i) robust for the skew in corpus sizes; (ii) capable of detecting semantic differences in infrequent words; and (iii) effective in pinpointing word instances that have a meaning missing in one of the two corpora under comparison. We show these advantages for historical corpora and also for native/non-native English corpora. | [
"Nagata, Ryo",
"Takamura, Hiroya",
"Otani, Naoki",
"Kawasaki, Yoshifumi"
] | Variance Matters: Detecting Semantic Differences without Corpus/Word Alignment | emnlp-main.965 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
||
https://aclanthology.org/2023.emnlp-main.966.bib | https://aclanthology.org/2023.emnlp-main.966/ | @inproceedings{liu-etal-2023-molca,
title = "{M}ol{CA}: Molecular Graph-Language Modeling with Cross-Modal Projector and Uni-Modal Adapter",
author = "Liu, Zhiyuan and
Li, Sihang and
Luo, Yanchen and
Fei, Hao and
Cao, Yixin and
Kawaguchi, Kenji and
Wang, Xiang and
Chua, Tat-Seng",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.966",
doi = "10.18653/v1/2023.emnlp-main.966",
pages = "15623--15638",
abstract = "Language Models (LMs) have demonstrated impressive molecule understanding ability on various 1D text-related tasks. However, they inherently lack 2D graph perception {---} a critical ability of human professionals in comprehending molecules{'} topological structures. To bridge this gap, we propose MolCA: Molecular Graph-Language Modeling with Cross-Modal Projector and Uni-Modal Adapter. MolCA enables an LM (i.e., Galactica) to understand both text- and graph-based molecular contents via the cross-modal projector. Specifically, the cross-modal projector is implemented as a Q-Former to connect a graph encoder{'}s representation space and an LM{'}s text space. Further, MolCA employs a uni-modal adapter (i.e., LoRA) for the LM{'}s efficient adaptation to downstream tasks. Unlike previous studies that couple an LM with a graph encoder via cross-modal contrastive learning, MolCA retains the LM{'}s ability of open-ended text generation and augments it with 2D graph information. To showcase its effectiveness, we extensively benchmark MolCA on tasks of molecule captioning, IUPAC name prediction, and molecule-text retrieval, on which MolCA significantly outperforms the baselines.",
}
| Language Models (LMs) have demonstrated impressive molecule understanding ability on various 1D text-related tasks. However, they inherently lack 2D graph perception {---} a critical ability of human professionals in comprehending molecules{'} topological structures. To bridge this gap, we propose MolCA: Molecular Graph-Language Modeling with Cross-Modal Projector and Uni-Modal Adapter. MolCA enables an LM (i.e., Galactica) to understand both text- and graph-based molecular contents via the cross-modal projector. Specifically, the cross-modal projector is implemented as a Q-Former to connect a graph encoder{'}s representation space and an LM{'}s text space. Further, MolCA employs a uni-modal adapter (i.e., LoRA) for the LM{'}s efficient adaptation to downstream tasks. Unlike previous studies that couple an LM with a graph encoder via cross-modal contrastive learning, MolCA retains the LM{'}s ability of open-ended text generation and augments it with 2D graph information. To showcase its effectiveness, we extensively benchmark MolCA on tasks of molecule captioning, IUPAC name prediction, and molecule-text retrieval, on which MolCA significantly outperforms the baselines. | [
"Liu, Zhiyuan",
"Li, Sihang",
"Luo, Yanchen",
"Fei, Hao",
"Cao, Yixin",
"Kawaguchi, Kenji",
"Wang, Xiang",
"Chua, Tat-Seng"
] | MolCA: Molecular Graph-Language Modeling with Cross-Modal Projector and Uni-Modal Adapter | emnlp-main.966 | 2310.12798 | [
"https://github.com/acharkq/molca"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.967.bib | https://aclanthology.org/2023.emnlp-main.967/ | @inproceedings{tu-etal-2023-training,
title = "A Training-Free Debiasing Framework with Counterfactual Reasoning for Conversational Emotion Detection",
author = "Tu, Geng and
Jing, Ran and
Liang, Bin and
Yang, Min and
Wong, Kam-Fai and
Xu, Ruifeng",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.967",
doi = "10.18653/v1/2023.emnlp-main.967",
pages = "15639--15650",
abstract = "Unintended dataset biases typically exist in existing Emotion Recognition in Conversations (ERC) datasets, including label bias, where models favor the majority class due to imbalanced training data, as well as the speaker and neutral word bias, where models make unfair predictions because of excessive correlations between specific neutral words or speakers and classes. However, previous studies in ERC generally focus on capturing context-sensitive and speaker-sensitive dependencies, ignoring the unintended dataset biases of data, which hampers the generalization and fairness in ERC. To address this issue, we propose a Training-Free Debiasing framework (TFD) that operates during prediction without additional training. To ensure compatibility with various ERC models, it does not balance data or modify the model structure. Instead, TFD extracts biases from the model by generating counterfactual utterances and contexts and mitigates them using simple yet empirically robust element-wise subtraction operations. Extensive experiments on three public datasets demonstrate that TFD effectively improves generalization ability and fairness across different ERC models.",
}
| Unintended dataset biases typically exist in existing Emotion Recognition in Conversations (ERC) datasets, including label bias, where models favor the majority class due to imbalanced training data, as well as the speaker and neutral word bias, where models make unfair predictions because of excessive correlations between specific neutral words or speakers and classes. However, previous studies in ERC generally focus on capturing context-sensitive and speaker-sensitive dependencies, ignoring the unintended dataset biases of data, which hampers the generalization and fairness in ERC. To address this issue, we propose a Training-Free Debiasing framework (TFD) that operates during prediction without additional training. To ensure compatibility with various ERC models, it does not balance data or modify the model structure. Instead, TFD extracts biases from the model by generating counterfactual utterances and contexts and mitigates them using simple yet empirically robust element-wise subtraction operations. Extensive experiments on three public datasets demonstrate that TFD effectively improves generalization ability and fairness across different ERC models. | [
"Tu, Geng",
"Jing, Ran",
"Liang, Bin",
"Yang, Min",
"Wong, Kam-Fai",
"Xu, Ruifeng"
] | A Training-Free Debiasing Framework with Counterfactual Reasoning for Conversational Emotion Detection | emnlp-main.967 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
||
https://aclanthology.org/2023.emnlp-main.968.bib | https://aclanthology.org/2023.emnlp-main.968/ | @inproceedings{chen-etal-2023-self,
title = "Self-{ICL}: Zero-Shot In-Context Learning with Self-Generated Demonstrations",
author = "Chen, Wei-Lin and
Wu, Cheng-Kuang and
Chen, Yun-Nung and
Chen, Hsin-Hsi",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.968",
doi = "10.18653/v1/2023.emnlp-main.968",
pages = "15651--15662",
abstract = "Large language models (LLMs) have exhibited striking in-context learning (ICL) ability to adapt to target tasks with a few input-output demonstrations. For better ICL, different methods are proposed to select representative demonstrations from existing training corpora. However, such settings are not aligned with real-world practices, as end-users usually query LMs without access to demonstration pools. In this work, we introduce Self-ICL{---}a simple framework which bootstraps LMs{'} intrinsic capabilities to perform zero-shot ICL. Given a test input, Self-ICL first prompts the model to generate pseudo-inputs. Next, the model predicts pseudo-labels for the pseudo-inputs via zero-shot prompting. Finally, we perform ICL for the test input with the pseudo-input-label pairs as demonstrations. Evaluation on 23 BIG-Bench Hard tasks shows Self-ICL outperforms zero-shot baselines on both average accuracy and head-to-head comparison. Moreover, with zero-shot chain-of-thought, Self-ICL achieves results comparable to using real demonstrations. Additionally, we conduct a range of analyses to validate Self-ICL{'}s effectiveness and provide insights for its behaviors under different settings.",
}
| Large language models (LLMs) have exhibited striking in-context learning (ICL) ability to adapt to target tasks with a few input-output demonstrations. For better ICL, different methods are proposed to select representative demonstrations from existing training corpora. However, such settings are not aligned with real-world practices, as end-users usually query LMs without access to demonstration pools. In this work, we introduce Self-ICL{---}a simple framework which bootstraps LMs{'} intrinsic capabilities to perform zero-shot ICL. Given a test input, Self-ICL first prompts the model to generate pseudo-inputs. Next, the model predicts pseudo-labels for the pseudo-inputs via zero-shot prompting. Finally, we perform ICL for the test input with the pseudo-input-label pairs as demonstrations. Evaluation on 23 BIG-Bench Hard tasks shows Self-ICL outperforms zero-shot baselines on both average accuracy and head-to-head comparison. Moreover, with zero-shot chain-of-thought, Self-ICL achieves results comparable to using real demonstrations. Additionally, we conduct a range of analyses to validate Self-ICL{'}s effectiveness and provide insights for its behaviors under different settings. | [
"Chen, Wei-Lin",
"Wu, Cheng-Kuang",
"Chen, Yun-Nung",
"Chen, Hsin-Hsi"
] | Self-ICL: Zero-Shot In-Context Learning with Self-Generated Demonstrations | emnlp-main.968 | 2305.15035 | [
"https://github.com/ntunlplab/self-icl"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.969.bib | https://aclanthology.org/2023.emnlp-main.969/ | @inproceedings{zhang-etal-2023-learning-knowledge,
title = "Learning Knowledge-Enhanced Contextual Language Representations for Domain Natural Language Understanding",
author = "Zhang, Taolin and
Xu, Ruyao and
Wang, Chengyu and
Duan, Zhongjie and
Chen, Cen and
Qiu, Minghui and
Cheng, Dawei and
He, Xiaofeng and
Qian, Weining",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.969",
doi = "10.18653/v1/2023.emnlp-main.969",
pages = "15663--15676",
abstract = "Knowledge-Enhanced Pre-trained Language Models (KEPLMs) improve the performance of various downstream NLP tasks by injecting knowledge facts from large-scale Knowledge Graphs (KGs). However, existing methods for pre-training KEPLMs with relational triples are difficult to be adapted to close domains due to the lack of sufficient domain graph semantics. In this paper, we propose a Knowledge-enhanced language representation learning framework for various closed domains (KANGAROO) via capturing the implicit graph structure among the entities. Specifically, since the entity coverage rates of closed-domain KGs can be relatively low and may exhibit the global sparsity phenomenon for knowledge injection, we consider not only the shallow relational representations of triples but also the hyperbolic embeddings of deep hierarchical entity-class structures for effective knowledge fusion. Moreover, as two closed-domain entities under the same entity-class often havel locally dense neighbor subgraphs counted by max point biconnected component, we further propose a data augmentation strategy based on contrastive learning over subgraphs to construct hard negative samples of higher quality. It makes the underlying KELPMs better distinguish the semantics of these neighboring entities to further complement the global semantic sparsity. In the experiments, we evaluate KANGAROO over various knowledge-aware and general NLP tasks in both full and few-shot learning settings, outperforming various KEPLM training paradigms performance in closed-domains significantly.",
}
| Knowledge-Enhanced Pre-trained Language Models (KEPLMs) improve the performance of various downstream NLP tasks by injecting knowledge facts from large-scale Knowledge Graphs (KGs). However, existing methods for pre-training KEPLMs with relational triples are difficult to be adapted to close domains due to the lack of sufficient domain graph semantics. In this paper, we propose a Knowledge-enhanced language representation learning framework for various closed domains (KANGAROO) via capturing the implicit graph structure among the entities. Specifically, since the entity coverage rates of closed-domain KGs can be relatively low and may exhibit the global sparsity phenomenon for knowledge injection, we consider not only the shallow relational representations of triples but also the hyperbolic embeddings of deep hierarchical entity-class structures for effective knowledge fusion. Moreover, as two closed-domain entities under the same entity-class often havel locally dense neighbor subgraphs counted by max point biconnected component, we further propose a data augmentation strategy based on contrastive learning over subgraphs to construct hard negative samples of higher quality. It makes the underlying KELPMs better distinguish the semantics of these neighboring entities to further complement the global semantic sparsity. In the experiments, we evaluate KANGAROO over various knowledge-aware and general NLP tasks in both full and few-shot learning settings, outperforming various KEPLM training paradigms performance in closed-domains significantly. | [
"Zhang, Taolin",
"Xu, Ruyao",
"Wang, Chengyu",
"Duan, Zhongjie",
"Chen, Cen",
"Qiu, Minghui",
"Cheng, Dawei",
"He, Xiaofeng",
"Qian, Weining"
] | Learning Knowledge-Enhanced Contextual Language Representations for Domain Natural Language Understanding | emnlp-main.969 | 2311.06761 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.970.bib | https://aclanthology.org/2023.emnlp-main.970/ | @inproceedings{wei-li-2023-scdner,
title = "{S}cd{NER}: Span-Based Consistency-Aware Document-Level Named Entity Recognition",
author = "Wei, Ying and
Li, Qi",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.970",
doi = "10.18653/v1/2023.emnlp-main.970",
pages = "15677--15685",
abstract = "Document-level NER approaches use global information via word-based key-value memory for accurate and consistent predictions. However, such global information on word level can introduce noise when the same word appears in different token sequences and has different labels. This work proposes a two-stage document-level NER model, ScdNER, for more accurate and consistent predictions via adaptive span-level global feature fusion. In the first stage, ScdNER trains a binary classifier to predict if a token sequence is an entity with a probability. Via a span-based key-value memory, the probabilities are further used to obtain the entity{'}s global features with reduced impact of non-entity sequences. The second stage predicts the entity types using a gate mechanism to balance its local and global information, leading to adaptive global feature fusion. Experiments on benchmark datasets from scientific, biomedical, and general domains show the effectiveness of the proposed methods.",
}
| Document-level NER approaches use global information via word-based key-value memory for accurate and consistent predictions. However, such global information on word level can introduce noise when the same word appears in different token sequences and has different labels. This work proposes a two-stage document-level NER model, ScdNER, for more accurate and consistent predictions via adaptive span-level global feature fusion. In the first stage, ScdNER trains a binary classifier to predict if a token sequence is an entity with a probability. Via a span-based key-value memory, the probabilities are further used to obtain the entity{'}s global features with reduced impact of non-entity sequences. The second stage predicts the entity types using a gate mechanism to balance its local and global information, leading to adaptive global feature fusion. Experiments on benchmark datasets from scientific, biomedical, and general domains show the effectiveness of the proposed methods. | [
"Wei, Ying",
"Li, Qi"
] | ScdNER: Span-Based Consistency-Aware Document-Level Named Entity Recognition | emnlp-main.970 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
||
https://aclanthology.org/2023.emnlp-main.971.bib | https://aclanthology.org/2023.emnlp-main.971/ | @inproceedings{zhong-etal-2023-mquake,
title = "{MQ}u{AKE}: Assessing Knowledge Editing in Language Models via Multi-Hop Questions",
author = "Zhong, Zexuan and
Wu, Zhengxuan and
Manning, Christopher and
Potts, Christopher and
Chen, Danqi",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.971",
doi = "10.18653/v1/2023.emnlp-main.971",
pages = "15686--15702",
abstract = "The information stored in large language models (LLMs) falls out of date quickly, and retraining from scratch is often not an option. This has recently given rise to a range of techniques for injecting new facts through updating model weights. Current evaluation paradigms are extremely limited, mainly validating the recall of edited facts, but changing one fact should cause rippling changes to the model{'}s related beliefs. If we edit the UK Prime Minister to now be Rishi Sunak, then we should get a different answer to Who is married to the British Prime Minister? In this work, we present a benchmark MQuAKE (Multi-hop Question Answering for Knowledge Editing) comprising multi-hop questions that assess whether edited models correctly answer questions where the answer should change as an entailed consequence of edited facts. While we find that current knowledge-editing approaches can recall edited facts accurately, they fail catastrophically on the constructed multi-hop questions. We thus propose a simple memory-based approach, MeLLo, which stores all edited facts externally while prompting the language model iteratively to generate answers that are consistent with the edited facts. While MQuAKE remains challenging, we show that MeLLo scales well with LLMs (up to 175B) and outperforms previous model editors by a large margin.",
}
| The information stored in large language models (LLMs) falls out of date quickly, and retraining from scratch is often not an option. This has recently given rise to a range of techniques for injecting new facts through updating model weights. Current evaluation paradigms are extremely limited, mainly validating the recall of edited facts, but changing one fact should cause rippling changes to the model{'}s related beliefs. If we edit the UK Prime Minister to now be Rishi Sunak, then we should get a different answer to Who is married to the British Prime Minister? In this work, we present a benchmark MQuAKE (Multi-hop Question Answering for Knowledge Editing) comprising multi-hop questions that assess whether edited models correctly answer questions where the answer should change as an entailed consequence of edited facts. While we find that current knowledge-editing approaches can recall edited facts accurately, they fail catastrophically on the constructed multi-hop questions. We thus propose a simple memory-based approach, MeLLo, which stores all edited facts externally while prompting the language model iteratively to generate answers that are consistent with the edited facts. While MQuAKE remains challenging, we show that MeLLo scales well with LLMs (up to 175B) and outperforms previous model editors by a large margin. | [
"Zhong, Zexuan",
"Wu, Zhengxuan",
"Manning, Christopher",
"Potts, Christopher",
"Chen, Danqi"
] | MQuAKE: Assessing Knowledge Editing in Language Models via Multi-Hop Questions | emnlp-main.971 | 2305.14795 | [
"https://github.com/princeton-nlp/mquake"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.972.bib | https://aclanthology.org/2023.emnlp-main.972/ | @inproceedings{li-etal-2023-stance,
title = "Stance Detection on Social Media with Background Knowledge",
author = "Li, Ang and
Liang, Bin and
Zhao, Jingqian and
Zhang, Bowen and
Yang, Min and
Xu, Ruifeng",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.972",
doi = "10.18653/v1/2023.emnlp-main.972",
pages = "15703--15717",
abstract = "Identifying users{'} stances regarding specific targets/topics is a significant route to learning public opinion from social media platforms. Most existing studies of stance detection strive to learn stance information about specific targets from the context, in order to determine the user{'}s stance on the target. However, in real-world scenarios, we usually have a certain understanding of a target when we express our stance on it. In this paper, we investigate stance detection from a novel perspective, where the background knowledge of the targets is taken into account for better stance detection. To be specific, we categorize background knowledge into two categories: episodic knowledge and discourse knowledge, and propose a novel Knowledge-Augmented Stance Detection (KASD) framework. For episodic knowledge, we devise a heuristic retrieval algorithm based on the topic to retrieve the Wikipedia documents relevant to the sample. Further, we construct a prompt for ChatGPT to filter the Wikipedia documents to derive episodic knowledge. For discourse knowledge, we construct a prompt for ChatGPT to paraphrase the hashtags, references, etc., in the sample, thereby injecting discourse knowledge into the sample. Experimental results on four benchmark datasets demonstrate that our KASD achieves state-of-the-art performance in in-target and zero-shot stance detection.",
}
| Identifying users{'} stances regarding specific targets/topics is a significant route to learning public opinion from social media platforms. Most existing studies of stance detection strive to learn stance information about specific targets from the context, in order to determine the user{'}s stance on the target. However, in real-world scenarios, we usually have a certain understanding of a target when we express our stance on it. In this paper, we investigate stance detection from a novel perspective, where the background knowledge of the targets is taken into account for better stance detection. To be specific, we categorize background knowledge into two categories: episodic knowledge and discourse knowledge, and propose a novel Knowledge-Augmented Stance Detection (KASD) framework. For episodic knowledge, we devise a heuristic retrieval algorithm based on the topic to retrieve the Wikipedia documents relevant to the sample. Further, we construct a prompt for ChatGPT to filter the Wikipedia documents to derive episodic knowledge. For discourse knowledge, we construct a prompt for ChatGPT to paraphrase the hashtags, references, etc., in the sample, thereby injecting discourse knowledge into the sample. Experimental results on four benchmark datasets demonstrate that our KASD achieves state-of-the-art performance in in-target and zero-shot stance detection. | [
"Li, Ang",
"Liang, Bin",
"Zhao, Jingqian",
"Zhang, Bowen",
"Yang, Min",
"Xu, Ruifeng"
] | Stance Detection on Social Media with Background Knowledge | emnlp-main.972 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
||
https://aclanthology.org/2023.emnlp-main.973.bib | https://aclanthology.org/2023.emnlp-main.973/ | @inproceedings{wang-etal-2023-vision,
title = "Vision-Enhanced Semantic Entity Recognition in Document Images via Visually-Asymmetric Consistency Learning",
author = "Wang, Hao and
Chen, Xiahua and
Wang, Rui and
Chu, Chenhui",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.973",
doi = "10.18653/v1/2023.emnlp-main.973",
pages = "15718--15731",
abstract = "Extracting meaningful entities belonging to predefined categories from Visually-rich Form-like Documents (VFDs) is a challenging task. Visual and layout features such as font, background, color, and bounding box location and size provide important cues for identifying entities of the same type. However, existing models commonly train a visual encoder with weak cross-modal supervision signals, resulting in a limited capacity to capture these non-textual features and suboptimal performance. In this paper, we propose a novel Visually-Asymmetric coNsistenCy Learning (VANCL) approach that addresses the above limitation by enhancing the model{'}s ability to capture fine-grained visual and layout features through the incorporation of color priors. Experimental results on benchmark datasets show that our approach substantially outperforms the strong LayoutLM series baseline, demonstrating the effectiveness of our approach. Additionally, we investigate the effects of different color schemes on our approach, providing insights for optimizing model performance. We believe our work will inspire future research on multimodal information extraction.",
}
| Extracting meaningful entities belonging to predefined categories from Visually-rich Form-like Documents (VFDs) is a challenging task. Visual and layout features such as font, background, color, and bounding box location and size provide important cues for identifying entities of the same type. However, existing models commonly train a visual encoder with weak cross-modal supervision signals, resulting in a limited capacity to capture these non-textual features and suboptimal performance. In this paper, we propose a novel Visually-Asymmetric coNsistenCy Learning (VANCL) approach that addresses the above limitation by enhancing the model{'}s ability to capture fine-grained visual and layout features through the incorporation of color priors. Experimental results on benchmark datasets show that our approach substantially outperforms the strong LayoutLM series baseline, demonstrating the effectiveness of our approach. Additionally, we investigate the effects of different color schemes on our approach, providing insights for optimizing model performance. We believe our work will inspire future research on multimodal information extraction. | [
"Wang, Hao",
"Chen, Xiahua",
"Wang, Rui",
"Chu, Chenhui"
] | Vision-Enhanced Semantic Entity Recognition in Document Images via Visually-Asymmetric Consistency Learning | emnlp-main.973 | 2310.14785 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.974.bib | https://aclanthology.org/2023.emnlp-main.974/ | @inproceedings{li-etal-2023-normdial,
title = "{N}orm{D}ial: A Comparable Bilingual Synthetic Dialog Dataset for Modeling Social Norm Adherence and Violation",
author = "Li, Oliver and
Subramanian, Mallika and
Saakyan, Arkadiy and
CH-Wang, Sky and
Muresan, Smaranda",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.974",
doi = "10.18653/v1/2023.emnlp-main.974",
pages = "15732--15744",
abstract = "Social norms fundamentally shape interpersonal communication. We present NormDial, a high-quality dyadic dialogue dataset with turn-by-turn annotations of social norm adherences and violations for Chinese and American cultures. Introducing the task of social norm observance detection, our dataset is synthetically generated in both Chinese and English using a human-in-the-loop pipeline by prompting large language models with a small collection of expert-annotated social norms. We show that our generated dialogues are of high quality through human evaluation and further evaluate the performance of existing large language models on this task. Our findings point towards new directions for understanding the nuances of social norms as they manifest in conversational contexts that span across languages and cultures.",
}
| Social norms fundamentally shape interpersonal communication. We present NormDial, a high-quality dyadic dialogue dataset with turn-by-turn annotations of social norm adherences and violations for Chinese and American cultures. Introducing the task of social norm observance detection, our dataset is synthetically generated in both Chinese and English using a human-in-the-loop pipeline by prompting large language models with a small collection of expert-annotated social norms. We show that our generated dialogues are of high quality through human evaluation and further evaluate the performance of existing large language models on this task. Our findings point towards new directions for understanding the nuances of social norms as they manifest in conversational contexts that span across languages and cultures. | [
"Li, Oliver",
"Subramanian, Mallika",
"Saakyan, Arkadiy",
"CH-Wang, Sky",
"Muresan, Smar",
"a"
] | NormDial: A Comparable Bilingual Synthetic Dialog Dataset for Modeling Social Norm Adherence and Violation | emnlp-main.974 | 2310.14563 | [
"https://github.com/aochong-li/normdial"
] | https://huggingface.co/papers/2310.14563 | 0 | 0 | 0 | 5 | [] | [] | [] | 1 | Poster |
https://aclanthology.org/2023.emnlp-main.975.bib | https://aclanthology.org/2023.emnlp-main.975/ | @inproceedings{schimanski-etal-2023-climatebert,
title = "{C}limate{BERT}-{N}et{Z}ero: Detecting and Assessing Net Zero and Reduction Targets",
author = "Schimanski, Tobias and
Bingler, Julia and
Kraus, Mathias and
Hyslop, Camilla and
Leippold, Markus",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.975",
doi = "10.18653/v1/2023.emnlp-main.975",
pages = "15745--15756",
abstract = "Public and private actors struggle to assess the vast amounts of information about sustainability commitments made by various institutions. To address this problem, we create a novel tool for automatically detecting corporate and national net zero and reduction targets in three steps. First, we introduce an expert-annotated data set with 3.5K text samples. Second, we train and release ClimateBERT-NetZero, a natural language classifier to detect whether a text contains a net zero or reduction target. Third, we showcase its analysis potential with two use cases: We first demonstrate how ClimateBERT-NetZero can be combined with conventional question-answering (Q{\&}A) models to analyze the ambitions displayed in net zero and reduction targets. Furthermore, we employ the ClimateBERT-NetZero model on quarterly earning call transcripts and outline how communication patterns evolve over time. Our experiments demonstrate promising pathways for extracting and analyzing net zero and emission reduction targets at scale.",
}
| Public and private actors struggle to assess the vast amounts of information about sustainability commitments made by various institutions. To address this problem, we create a novel tool for automatically detecting corporate and national net zero and reduction targets in three steps. First, we introduce an expert-annotated data set with 3.5K text samples. Second, we train and release ClimateBERT-NetZero, a natural language classifier to detect whether a text contains a net zero or reduction target. Third, we showcase its analysis potential with two use cases: We first demonstrate how ClimateBERT-NetZero can be combined with conventional question-answering (Q{\&}A) models to analyze the ambitions displayed in net zero and reduction targets. Furthermore, we employ the ClimateBERT-NetZero model on quarterly earning call transcripts and outline how communication patterns evolve over time. Our experiments demonstrate promising pathways for extracting and analyzing net zero and emission reduction targets at scale. | [
"Schimanski, Tobias",
"Bingler, Julia",
"Kraus, Mathias",
"Hyslop, Camilla",
"Leippold, Markus"
] | ClimateBERT-NetZero: Detecting and Assessing Net Zero and Reduction Targets | emnlp-main.975 | 2310.08096 | [
""
] | https://huggingface.co/papers/2310.08096 | 0 | 0 | 0 | 5 | [
"climatebert/netzero-reduction"
] | [] | [] | 1 | Oral |
https://aclanthology.org/2023.emnlp-main.976.bib | https://aclanthology.org/2023.emnlp-main.976/ | @inproceedings{kim-etal-2023-leap,
title = "Leap-of-Thought: Accelerating Transformers via Dynamic Token Routing",
author = "Kim, Yeachan and
Kim, Junho and
Park, Jun-Hyung and
Lee, Mingyu and
Lee, SangKeun",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.976",
doi = "10.18653/v1/2023.emnlp-main.976",
pages = "15757--15769",
abstract = "Computational inefficiency in transformers has been a long-standing challenge, hindering the deployment in resource-constrained or real-time applications. One promising approach to mitigate this limitation is to progressively remove less significant tokens, given that the sequence length strongly contributes to the inefficiency. However, this approach entails a potential risk of losing crucial information due to the irrevocable nature of token removal. In this paper, we introduce Leap-of-Thought (LoT), a novel token reduction approach that dynamically routes tokens within layers. Unlike previous work that irrevocably discards tokens, LoT enables tokens to {`}leap{'} across layers. This ensures that all tokens remain accessible in subsequent layers while reducing the number of tokens processed within layers. We achieve this by pairing the transformer with dynamic token routers, which learn to selectively process tokens essential for the task. Evaluation results clearly show that LoT achieves a substantial improvement in computational efficiency. Specifically, LoT attains up to 25x faster inference time without a significant loss in accuracy",
}
| Computational inefficiency in transformers has been a long-standing challenge, hindering the deployment in resource-constrained or real-time applications. One promising approach to mitigate this limitation is to progressively remove less significant tokens, given that the sequence length strongly contributes to the inefficiency. However, this approach entails a potential risk of losing crucial information due to the irrevocable nature of token removal. In this paper, we introduce Leap-of-Thought (LoT), a novel token reduction approach that dynamically routes tokens within layers. Unlike previous work that irrevocably discards tokens, LoT enables tokens to {`}leap{'} across layers. This ensures that all tokens remain accessible in subsequent layers while reducing the number of tokens processed within layers. We achieve this by pairing the transformer with dynamic token routers, which learn to selectively process tokens essential for the task. Evaluation results clearly show that LoT achieves a substantial improvement in computational efficiency. Specifically, LoT attains up to 25x faster inference time without a significant loss in accuracy | [
"Kim, Yeachan",
"Kim, Junho",
"Park, Jun-Hyung",
"Lee, Mingyu",
"Lee, SangKeun"
] | Leap-of-Thought: Accelerating Transformers via Dynamic Token Routing | emnlp-main.976 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
||
https://aclanthology.org/2023.emnlp-main.977.bib | https://aclanthology.org/2023.emnlp-main.977/ | @inproceedings{nath-etal-2023-reinforcement,
title = "Reinforcement Replaces Supervision: Query focused Summarization using Deep Reinforcement Learning",
author = "Nath, Swaroop and
Bhattacharyya, Pushpak and
Khadilkar, Harshad",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.977",
doi = "10.18653/v1/2023.emnlp-main.977",
pages = "15770--15789",
abstract = "Query-focused Summarization (QfS) deals with systems that generate summaries from document(s) based on a query. Motivated by the insight that Reinforcement Learning (RL) provides a generalization to Supervised Learning (SL) for Natural Language Generation, and thereby performs better (empirically) than SL, we use an RL-based approach for this task of QfS. Additionally, we also resolve the conflict of employing RL in Transformers with Teacher Forcing. We develop multiple Policy Gradient networks, trained on various reward signals: ROUGE, BLEU, and Semantic Similarity, which lead to a $\mathit{10}$-point improvement over the State-of-the-Art approach on the ROUGE-L metric for a benchmark dataset (ELI5). We also show performance of our approach in zero-shot setting for another benchmark dataset (DebatePedia) {--} our approach leads to results comparable to baselines, which were specifically trained on DebatePedia. To aid the RL training, we propose a better semantic similarity reward, enabled by a novel Passage Embedding scheme developed using Cluster Hypothesis. Lastly, we contribute a gold-standard test dataset to further research in QfS and Long-form Question Answering (LfQA).",
}
| Query-focused Summarization (QfS) deals with systems that generate summaries from document(s) based on a query. Motivated by the insight that Reinforcement Learning (RL) provides a generalization to Supervised Learning (SL) for Natural Language Generation, and thereby performs better (empirically) than SL, we use an RL-based approach for this task of QfS. Additionally, we also resolve the conflict of employing RL in Transformers with Teacher Forcing. We develop multiple Policy Gradient networks, trained on various reward signals: ROUGE, BLEU, and Semantic Similarity, which lead to a $\mathit{10}$-point improvement over the State-of-the-Art approach on the ROUGE-L metric for a benchmark dataset (ELI5). We also show performance of our approach in zero-shot setting for another benchmark dataset (DebatePedia) {--} our approach leads to results comparable to baselines, which were specifically trained on DebatePedia. To aid the RL training, we propose a better semantic similarity reward, enabled by a novel Passage Embedding scheme developed using Cluster Hypothesis. Lastly, we contribute a gold-standard test dataset to further research in QfS and Long-form Question Answering (LfQA). | [
"Nath, Swaroop",
"Bhattacharyya, Pushpak",
"Khadilkar, Harshad"
] | Reinforcement Replaces Supervision: Query focused Summarization using Deep Reinforcement Learning | emnlp-main.977 | 2311.17514 | [
"https://github.com/swaroop-nath/rl-qfs"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.978.bib | https://aclanthology.org/2023.emnlp-main.978/ | @inproceedings{leteno-etal-2023-fair,
title = "Fair Text Classification with {W}asserstein Independence",
author = "Leteno, Thibaud and
Gourru, Antoine and
Laclau, Charlotte and
Emonet, R{\'e}mi and
Gravier, Christophe",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.978",
doi = "10.18653/v1/2023.emnlp-main.978",
pages = "15790--15803",
abstract = "Group fairness is a central research topic in text classification, where reaching fair treatment between sensitive groups (e.g. women vs. men) remains an open challenge. This paper presents a novel method for mitigating biases in neural text classification, agnostic to the model architecture. Considering the difficulty to distinguish fair from unfair information in a text encoder, we take inspiration from adversarial training to induce Wasserstein independence between representations learned to predict our target label and the ones learned to predict some sensitive attribute. Our approach provides two significant advantages. Firstly, it does not require annotations of sensitive attributes in both testing and training data. This is more suitable for real-life scenarios compared to existing methods that require annotations of sensitive attributes at train time. Secondly, our approach exhibits a comparable or better fairness-accuracy trade-off compared to existing methods.",
}
| Group fairness is a central research topic in text classification, where reaching fair treatment between sensitive groups (e.g. women vs. men) remains an open challenge. This paper presents a novel method for mitigating biases in neural text classification, agnostic to the model architecture. Considering the difficulty to distinguish fair from unfair information in a text encoder, we take inspiration from adversarial training to induce Wasserstein independence between representations learned to predict our target label and the ones learned to predict some sensitive attribute. Our approach provides two significant advantages. Firstly, it does not require annotations of sensitive attributes in both testing and training data. This is more suitable for real-life scenarios compared to existing methods that require annotations of sensitive attributes at train time. Secondly, our approach exhibits a comparable or better fairness-accuracy trade-off compared to existing methods. | [
"Leteno, Thibaud",
"Gourru, Antoine",
"Laclau, Charlotte",
"Emonet, R{\\'e}mi",
"Gravier, Christophe"
] | Fair Text Classification with Wasserstein Independence | emnlp-main.978 | 2311.12689 | [
"https://github.com/letenothibaud/wasserstein_fair_classification"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Oral |
|
https://aclanthology.org/2023.emnlp-main.979.bib | https://aclanthology.org/2023.emnlp-main.979/ | @inproceedings{xu-etal-2023-tacoprompt,
title = "{T}aco{P}rompt: A Collaborative Multi-Task Prompt Learning Method for Self-Supervised Taxonomy Completion",
author = "Xu, Hongyuan and
Liu, Ciyi and
Niu, Yuhang and
Chen, Yunong and
Cai, Xiangrui and
Wen, Yanlong and
Yuan, Xiaojie",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.979",
doi = "10.18653/v1/2023.emnlp-main.979",
pages = "15804--15817",
abstract = "Automatic taxonomy completion aims to attach the emerging concept to an appropriate pair of hypernym and hyponym in the existing taxonomy. Existing methods suffer from the overfitting to leaf-only problem caused by imbalanced leaf and non-leaf samples when training the newly initialized classification head. Besides, they only leverage subtasks, namely attaching the concept to its hypernym or hyponym, as auxiliary supervision for representation learning yet neglect the effects of subtask results on the final prediction. To address the aforementioned limitations, we propose TacoPrompt, a Collaborative Multi-Task Prompt Learning Method for Self-Supervised Taxonomy Completion. First, we perform triplet semantic matching using the prompt learning paradigm to effectively learn non-leaf attachment ability from imbalanced training samples. Second, we design the result context to relate the final prediction to the subtask results by a contextual approach, enhancing prompt-based multi-task learning. Third, we leverage a two-stage retrieval and re-ranking approach to improve the inference efficiency. Experimental results on three datasets show that TacoPrompt achieves state-of-the-art taxonomy completion performance. Codes are available at https://github.com/cyclexu/TacoPrompt.",
}
| Automatic taxonomy completion aims to attach the emerging concept to an appropriate pair of hypernym and hyponym in the existing taxonomy. Existing methods suffer from the overfitting to leaf-only problem caused by imbalanced leaf and non-leaf samples when training the newly initialized classification head. Besides, they only leverage subtasks, namely attaching the concept to its hypernym or hyponym, as auxiliary supervision for representation learning yet neglect the effects of subtask results on the final prediction. To address the aforementioned limitations, we propose TacoPrompt, a Collaborative Multi-Task Prompt Learning Method for Self-Supervised Taxonomy Completion. First, we perform triplet semantic matching using the prompt learning paradigm to effectively learn non-leaf attachment ability from imbalanced training samples. Second, we design the result context to relate the final prediction to the subtask results by a contextual approach, enhancing prompt-based multi-task learning. Third, we leverage a two-stage retrieval and re-ranking approach to improve the inference efficiency. Experimental results on three datasets show that TacoPrompt achieves state-of-the-art taxonomy completion performance. Codes are available at https://github.com/cyclexu/TacoPrompt. | [
"Xu, Hongyuan",
"Liu, Ciyi",
"Niu, Yuhang",
"Chen, Yunong",
"Cai, Xiangrui",
"Wen, Yanlong",
"Yuan, Xiaojie"
] | TacoPrompt: A Collaborative Multi-Task Prompt Learning Method for Self-Supervised Taxonomy Completion | emnlp-main.979 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
||
https://aclanthology.org/2023.emnlp-main.980.bib | https://aclanthology.org/2023.emnlp-main.980/ | @inproceedings{moeller-etal-2023-attribution,
title = "An Attribution Method for {S}iamese Encoders",
author = "Moeller, Lucas and
Nikolaev, Dmitry and
Pad{\'o}, Sebastian",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.980",
doi = "10.18653/v1/2023.emnlp-main.980",
pages = "15818--15827",
abstract = "Despite the success of Siamese encoder models such as sentence transformers (ST), little is known about the aspects of inputs they pay attention to. A barrier is that their predictions cannot be attributed to individual features, as they compare two inputs rather than processing a single one. This paper derives a local attribution method for Siamese encoders by generalizing the principle of integrated gradients to models with multiple inputs. The output takes the form of feature-pair attributions and in case of STs it can be reduced to a token{--}token matrix. Our method involves the introduction of integrated Jacobians and inherits the advantageous formal properties of integrated gradients: it accounts for the model{'}s full computation graph and is guaranteed to converge to the actual prediction. A pilot study shows that in case of STs few token pairs can dominate predictions and that STs preferentially focus on nouns and verbs. For accurate predictions, however, they need to attend to the majority of tokens and parts of speech.",
}
| Despite the success of Siamese encoder models such as sentence transformers (ST), little is known about the aspects of inputs they pay attention to. A barrier is that their predictions cannot be attributed to individual features, as they compare two inputs rather than processing a single one. This paper derives a local attribution method for Siamese encoders by generalizing the principle of integrated gradients to models with multiple inputs. The output takes the form of feature-pair attributions and in case of STs it can be reduced to a token{--}token matrix. Our method involves the introduction of integrated Jacobians and inherits the advantageous formal properties of integrated gradients: it accounts for the model{'}s full computation graph and is guaranteed to converge to the actual prediction. A pilot study shows that in case of STs few token pairs can dominate predictions and that STs preferentially focus on nouns and verbs. For accurate predictions, however, they need to attend to the majority of tokens and parts of speech. | [
"Moeller, Lucas",
"Nikolaev, Dmitry",
"Pad{\\'o}, Sebastian"
] | An Attribution Method for Siamese Encoders | emnlp-main.980 | 2310.05703 | [
"https://github.com/lucasmllr/xsbert"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.981.bib | https://aclanthology.org/2023.emnlp-main.981/ | @inproceedings{mukherjee-etal-2023-global,
title = "{G}lobal {V}oices, Local Biases: Socio-Cultural Prejudices across Languages",
author = "Mukherjee, Anjishnu and
Raj, Chahat and
Zhu, Ziwei and
Anastasopoulos, Antonios",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.981",
doi = "10.18653/v1/2023.emnlp-main.981",
pages = "15828--15845",
abstract = "Human biases are ubiquitous but not uniform: disparities exist across linguistic, cultural, and societal borders. As large amounts of recent literature suggest, language models (LMs) trained on human data can reflect and often amplify the effects of these social biases. However, the vast majority of existing studies on bias are heavily skewed towards Western and European languages. In this work, we scale the Word Embedding Association Test (WEAT) to 24 languages, enabling broader studies and yielding interesting findings about LM bias. We additionally enhance this data with culturally relevant information for each language, capturing local contexts on a global scale. Further, to encompass more widely prevalent societal biases, we examine new bias dimensions across toxicity, ableism, and more. Moreover, we delve deeper into the Indian linguistic landscape, conducting a comprehensive regional bias analysis across six prevalent Indian languages. Finally, we highlight the significance of these social biases and the new dimensions through an extensive comparison of embedding methods, reinforcing the need to address them in pursuit of more equitable language models.",
}
| Human biases are ubiquitous but not uniform: disparities exist across linguistic, cultural, and societal borders. As large amounts of recent literature suggest, language models (LMs) trained on human data can reflect and often amplify the effects of these social biases. However, the vast majority of existing studies on bias are heavily skewed towards Western and European languages. In this work, we scale the Word Embedding Association Test (WEAT) to 24 languages, enabling broader studies and yielding interesting findings about LM bias. We additionally enhance this data with culturally relevant information for each language, capturing local contexts on a global scale. Further, to encompass more widely prevalent societal biases, we examine new bias dimensions across toxicity, ableism, and more. Moreover, we delve deeper into the Indian linguistic landscape, conducting a comprehensive regional bias analysis across six prevalent Indian languages. Finally, we highlight the significance of these social biases and the new dimensions through an extensive comparison of embedding methods, reinforcing the need to address them in pursuit of more equitable language models. | [
"Mukherjee, Anjishnu",
"Raj, Chahat",
"Zhu, Ziwei",
"Anastasopoulos, Antonios"
] | Global Voices, Local Biases: Socio-Cultural Prejudices across Languages | emnlp-main.981 | 2310.17586 | [
"https://github.com/iamshnoo/weathub"
] | https://huggingface.co/papers/2310.17586 | 0 | 0 | 0 | 4 | [] | [
"iamshnoo/WEATHub"
] | [] | 1 | Poster |
https://aclanthology.org/2023.emnlp-main.982.bib | https://aclanthology.org/2023.emnlp-main.982/ | @inproceedings{yang-etal-2023-graph,
title = "Graph vs. Sequence: An Empirical Study on Knowledge Forms for Knowledge-Grounded Dialogue",
author = "Yang, Yizhe and
Huang, Heyan and
Liu, Yuhang and
Gao, Yang",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.982",
doi = "10.18653/v1/2023.emnlp-main.982",
pages = "15846--15858",
abstract = "Knowledge-grounded dialogue is a task of gener- ating an informative response based on both the dialogue history and external knowledge source. In general, there are two forms of knowledge: manu- ally annotated knowledge graphs and knowledge text from website. From various evaluation viewpoints, each type of knowledge has advantages and downsides. To further distinguish the principles and determinants from the intricate factors, we conduct a thorough experiment and study on the task to answer three essential questions. The ques- tions involve the choice of appropriate knowledge form, the degree of mutual effects between knowl- edge and the model selection, and the few-shot performance of knowledge. Supported by statistical shreds of evidence, we offer conclusive solutions and sensible suggestions for directions and standards of future research.",
}
| Knowledge-grounded dialogue is a task of gener- ating an informative response based on both the dialogue history and external knowledge source. In general, there are two forms of knowledge: manu- ally annotated knowledge graphs and knowledge text from website. From various evaluation viewpoints, each type of knowledge has advantages and downsides. To further distinguish the principles and determinants from the intricate factors, we conduct a thorough experiment and study on the task to answer three essential questions. The ques- tions involve the choice of appropriate knowledge form, the degree of mutual effects between knowl- edge and the model selection, and the few-shot performance of knowledge. Supported by statistical shreds of evidence, we offer conclusive solutions and sensible suggestions for directions and standards of future research. | [
"Yang, Yizhe",
"Huang, Heyan",
"Liu, Yuhang",
"Gao, Yang"
] | Graph vs. Sequence: An Empirical Study on Knowledge Forms for Knowledge-Grounded Dialogue | emnlp-main.982 | 2312.07868 | [
""
] | https://huggingface.co/papers/2312.07868 | 1 | 0 | 0 | 4 | [] | [] | [] | 1 | Poster |
https://aclanthology.org/2023.emnlp-main.983.bib | https://aclanthology.org/2023.emnlp-main.983/ | @inproceedings{gee-etal-2023-compressed,
title = "Are Compressed Language Models Less Subgroup Robust?",
author = "Gee, Leonidas and
Zugarini, Andrea and
Quadrianto, Novi",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.983",
doi = "10.18653/v1/2023.emnlp-main.983",
pages = "15859--15868",
abstract = "To reduce the inference cost of large language models, model compression is increasingly used to create smaller scalable models. However, little is known about their robustness to minority subgroups defined by the labels and attributes of a dataset. In this paper, we investigate the effects of 18 different compression methods and settings on the subgroup robustness of BERT language models. We show that worst-group performance does not depend on model size alone, but also on the compression method used. Additionally, we find that model compression does not always worsen the performance on minority subgroups. Altogether, our analysis serves to further research into the subgroup robustness of model compression.",
}
| To reduce the inference cost of large language models, model compression is increasingly used to create smaller scalable models. However, little is known about their robustness to minority subgroups defined by the labels and attributes of a dataset. In this paper, we investigate the effects of 18 different compression methods and settings on the subgroup robustness of BERT language models. We show that worst-group performance does not depend on model size alone, but also on the compression method used. Additionally, we find that model compression does not always worsen the performance on minority subgroups. Altogether, our analysis serves to further research into the subgroup robustness of model compression. | [
"Gee, Leonidas",
"Zugarini, Andrea",
"Quadrianto, Novi"
] | Are Compressed Language Models Less Subgroup Robust? | emnlp-main.983 | 2403.17811 | [
"https://github.com/wearepal/compression-subgroup"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.984.bib | https://aclanthology.org/2023.emnlp-main.984/ | @inproceedings{guo-vosoughi-2023-length,
title = "Length Does Matter: Summary Length can Bias Summarization Metrics",
author = "Guo, Xiaobo and
Vosoughi, Soroush",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.984",
doi = "10.18653/v1/2023.emnlp-main.984",
pages = "15869--15879",
abstract = "Establishing the characteristics of an effective summary is a complicated and often subjective endeavor. Consequently, the development of metrics for the summarization task has become a dynamic area of research within natural language processing. In this paper, we reveal that existing summarization metrics exhibit a bias toward the length of generated summaries. Our thorough experiments, conducted on a variety of datasets, metrics, and models, substantiate these findings. The results indicate that most metrics tend to favor longer summaries, even after accounting for other factors. To address this issue, we introduce a Bayesian normalization technique that effectively diminishes this bias. We demonstrate that our approach significantly improves the concordance between human annotators and the majority of metrics in terms of summary coherence.",
}
| Establishing the characteristics of an effective summary is a complicated and often subjective endeavor. Consequently, the development of metrics for the summarization task has become a dynamic area of research within natural language processing. In this paper, we reveal that existing summarization metrics exhibit a bias toward the length of generated summaries. Our thorough experiments, conducted on a variety of datasets, metrics, and models, substantiate these findings. The results indicate that most metrics tend to favor longer summaries, even after accounting for other factors. To address this issue, we introduce a Bayesian normalization technique that effectively diminishes this bias. We demonstrate that our approach significantly improves the concordance between human annotators and the majority of metrics in terms of summary coherence. | [
"Guo, Xiaobo",
"Vosoughi, Soroush"
] | Length Does Matter: Summary Length can Bias Summarization Metrics | emnlp-main.984 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
||
https://aclanthology.org/2023.emnlp-main.985.bib | https://aclanthology.org/2023.emnlp-main.985/ | @inproceedings{chen-etal-2023-nl2tl,
title = "{NL}2{TL}: Transforming Natural Languages to Temporal Logics using Large Language Models",
author = "Chen, Yongchao and
Gandhi, Rujul and
Zhang, Yang and
Fan, Chuchu",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.985",
doi = "10.18653/v1/2023.emnlp-main.985",
pages = "15880--15903",
abstract = "Temporal Logic (TL) can be used to rigorously specify complex high-level specification for systems in many engineering applications. The translation between natural language (NL) and TL has been under-explored due to the lack of dataset and generalizable model across different application domains. In this paper, we propose an accurate and generalizable transformation framework of English instructions from NL to TL, exploring the use of Large Language Models (LLMs) at multiple stages. Our contributions are twofold. First, we develop a framework to create a dataset of NL-TL pairs combining LLMs and human annotation. We publish a dataset with 23K NL-TL pairs. Then, we finetune T5 models on the lifted versions (i.e., the specific Atomic Propositions (AP) are hidden) of the NL and TL. The enhanced generalizability originates from two aspects: 1) Usage of lifted NL-TL characterizes common logical structures, without constraints of specific domains. 2) Application of LLMs in dataset creation largely enhances corpus richness. We test the generalization of trained models on five varied domains. To achieve full NL-TL transformation, we either combine the lifted model with AP recognition task or do the further finetuning on each specific domain. During the further finetuning, our model achieves higher accuracy ({\textgreater} 95{\%}) using only {\textless}10{\%} training data, compared with the baseline sequence to sequence (Seq2Seq) model.",
}
| Temporal Logic (TL) can be used to rigorously specify complex high-level specification for systems in many engineering applications. The translation between natural language (NL) and TL has been under-explored due to the lack of dataset and generalizable model across different application domains. In this paper, we propose an accurate and generalizable transformation framework of English instructions from NL to TL, exploring the use of Large Language Models (LLMs) at multiple stages. Our contributions are twofold. First, we develop a framework to create a dataset of NL-TL pairs combining LLMs and human annotation. We publish a dataset with 23K NL-TL pairs. Then, we finetune T5 models on the lifted versions (i.e., the specific Atomic Propositions (AP) are hidden) of the NL and TL. The enhanced generalizability originates from two aspects: 1) Usage of lifted NL-TL characterizes common logical structures, without constraints of specific domains. 2) Application of LLMs in dataset creation largely enhances corpus richness. We test the generalization of trained models on five varied domains. To achieve full NL-TL transformation, we either combine the lifted model with AP recognition task or do the further finetuning on each specific domain. During the further finetuning, our model achieves higher accuracy ({\textgreater} 95{\%}) using only {\textless}10{\%} training data, compared with the baseline sequence to sequence (Seq2Seq) model. | [
"Chen, Yongchao",
"G",
"hi, Rujul",
"Zhang, Yang",
"Fan, Chuchu"
] | NL2TL: Transforming Natural Languages to Temporal Logics using Large Language Models | emnlp-main.985 | 2305.07766 | [
"https://github.com/yongchao98/nl2tl"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.986.bib | https://aclanthology.org/2023.emnlp-main.986/ | @inproceedings{gkoumas-etal-2023-reformulating,
title = "Reformulating {NLP} tasks to Capture Longitudinal Manifestation of Language Disorders in People with Dementia.",
author = "Gkoumas, Dimitris and
Purver, Matthew and
Liakata, Maria",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.986",
doi = "10.18653/v1/2023.emnlp-main.986",
pages = "15904--15917",
abstract = "Dementia is associated with language disorders which impede communication. Here, we automatically learn linguistic disorder patterns by making use of a moderately-sized pre-trained language model and forcing it to focus on reformulated natural language processing (NLP) tasks and associated linguistic patterns. Our experiments show that NLP tasks that encapsulate contextual information and enhance the gradient signal with linguistic patterns benefit performance. We then use the probability estimates from the best model to construct digital linguistic markers measuring the overall quality in communication and the intensity of a variety of language disorders. We investigate how the digital markers characterize dementia speech from a longitudinal perspective. We find that our proposed communication marker is able to robustly and reliably characterize the language of people with dementia, outperforming existing linguistic approaches; and shows external validity via significant correlation with clinical markers of behaviour. Finally, our proposed linguistic disorder markers provide useful insights into gradual language impairment associated with disease progression.",
}
| Dementia is associated with language disorders which impede communication. Here, we automatically learn linguistic disorder patterns by making use of a moderately-sized pre-trained language model and forcing it to focus on reformulated natural language processing (NLP) tasks and associated linguistic patterns. Our experiments show that NLP tasks that encapsulate contextual information and enhance the gradient signal with linguistic patterns benefit performance. We then use the probability estimates from the best model to construct digital linguistic markers measuring the overall quality in communication and the intensity of a variety of language disorders. We investigate how the digital markers characterize dementia speech from a longitudinal perspective. We find that our proposed communication marker is able to robustly and reliably characterize the language of people with dementia, outperforming existing linguistic approaches; and shows external validity via significant correlation with clinical markers of behaviour. Finally, our proposed linguistic disorder markers provide useful insights into gradual language impairment associated with disease progression. | [
"Gkoumas, Dimitris",
"Purver, Matthew",
"Liakata, Maria"
] | Reformulating NLP tasks to Capture Longitudinal Manifestation of Language Disorders in People with Dementia. | emnlp-main.986 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
||
https://aclanthology.org/2023.emnlp-main.987.bib | https://aclanthology.org/2023.emnlp-main.987/ | @inproceedings{mamta-etal-2023-elevating,
title = "Elevating Code-mixed Text Handling through Auditory Information of Words",
author = "Mamta, Mamta and
Ahmad, Zishan and
Ekbal, Asif",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.987",
doi = "10.18653/v1/2023.emnlp-main.987",
pages = "15918--15932",
abstract = "With the growing popularity of code-mixed data, there is an increasing need for better handling of this type of data, which poses a number of challenges, such as dealing with spelling variations, multiple languages, different scripts, and a lack of resources. Current language models face difficulty in effectively handling code-mixed data as they primarily focus on the semantic representation of words and ignore the auditory phonetic features. This leads to difficulties in handling spelling variations in code-mixed text. In this paper, we propose an effective approach for creating language models for handling code-mixed textual data using auditory information of words from SOUNDEX. Our approach includes a pre-training step based on masked-language-modelling, which includes SOUNDEX representations (SAMLM) and a new method of providing input data to the pre-trained model. Through experimentation on various code-mixed datasets (of different languages) for sentiment, offensive and aggression classification tasks, we establish that our novel language modeling approach (SAMLM) results in improved robustness towards adversarial attacks on code-mixed classification tasks. Additionally, our SAMLM based approach also results in better classification results over the popular baselines for code-mixed tasks. We use the explainability technique, SHAP (SHapley Additive exPlanations) to explain how the auditory features incorporated through SAMLM assist the model to handle the code-mixed text effectively and increase robustness against adversarial attacks.",
}
| With the growing popularity of code-mixed data, there is an increasing need for better handling of this type of data, which poses a number of challenges, such as dealing with spelling variations, multiple languages, different scripts, and a lack of resources. Current language models face difficulty in effectively handling code-mixed data as they primarily focus on the semantic representation of words and ignore the auditory phonetic features. This leads to difficulties in handling spelling variations in code-mixed text. In this paper, we propose an effective approach for creating language models for handling code-mixed textual data using auditory information of words from SOUNDEX. Our approach includes a pre-training step based on masked-language-modelling, which includes SOUNDEX representations (SAMLM) and a new method of providing input data to the pre-trained model. Through experimentation on various code-mixed datasets (of different languages) for sentiment, offensive and aggression classification tasks, we establish that our novel language modeling approach (SAMLM) results in improved robustness towards adversarial attacks on code-mixed classification tasks. Additionally, our SAMLM based approach also results in better classification results over the popular baselines for code-mixed tasks. We use the explainability technique, SHAP (SHapley Additive exPlanations) to explain how the auditory features incorporated through SAMLM assist the model to handle the code-mixed text effectively and increase robustness against adversarial attacks. | [
"Mamta, Mamta",
"Ahmad, Zishan",
"Ekbal, Asif"
] | Elevating Code-mixed Text Handling through Auditory Information of Words | emnlp-main.987 | 2310.18155 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.988.bib | https://aclanthology.org/2023.emnlp-main.988/ | @inproceedings{tiwari-etal-2023-predict,
title = "Predict and Use: Harnessing Predicted Gaze to Improve Multimodal Sarcasm Detection",
author = "Tiwari, Divyank and
Kanojia, Diptesh and
Ray, Anupama and
Nunna, Apoorva and
Bhattacharyya, Pushpak",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.988",
doi = "10.18653/v1/2023.emnlp-main.988",
pages = "15933--15948",
abstract = "Sarcasm is a complex linguistic construct with incongruity at its very core. Detecting sarcasm depends on the actual content spoken and tonality, facial expressions, the context of an utterance, and personal traits like language proficiency and cognitive capabilities. In this paper, we propose the utilization of synthetic gaze data to improve the task performance for multimodal sarcasm detection in a conversational setting. We enrich an existing multimodal conversational dataset, i.e., MUStARD++ with gaze features. With the help of human participants, we collect gaze features for 20{\%} of data instances, and we investigate various methods for gaze feature prediction for the rest of the dataset. We perform extrinsic and intrinsic evaluations to assess the quality of the predicted gaze features. We observe a performance gain of up to 6.6{\%} points by adding a new modality, i.e., collected gaze features. When both collected and predicted data are used, we observe a performance gain of 2.3{\%} points on the complete dataset. Interestingly, with only predicted gaze features, too, we observe a gain in performance (1.9{\%} points). We retain and use the feature prediction model, which maximally correlates with collected gaze features. Our model trained on combining collected and synthetic gaze data achieves SoTA performance on the MUStARD++ dataset. To the best of our knowledge, ours is the first predict-and-use model for sarcasm detection. We publicly release the code, gaze data, and our best models for further research.",
}
| Sarcasm is a complex linguistic construct with incongruity at its very core. Detecting sarcasm depends on the actual content spoken and tonality, facial expressions, the context of an utterance, and personal traits like language proficiency and cognitive capabilities. In this paper, we propose the utilization of synthetic gaze data to improve the task performance for multimodal sarcasm detection in a conversational setting. We enrich an existing multimodal conversational dataset, i.e., MUStARD++ with gaze features. With the help of human participants, we collect gaze features for 20{\%} of data instances, and we investigate various methods for gaze feature prediction for the rest of the dataset. We perform extrinsic and intrinsic evaluations to assess the quality of the predicted gaze features. We observe a performance gain of up to 6.6{\%} points by adding a new modality, i.e., collected gaze features. When both collected and predicted data are used, we observe a performance gain of 2.3{\%} points on the complete dataset. Interestingly, with only predicted gaze features, too, we observe a gain in performance (1.9{\%} points). We retain and use the feature prediction model, which maximally correlates with collected gaze features. Our model trained on combining collected and synthetic gaze data achieves SoTA performance on the MUStARD++ dataset. To the best of our knowledge, ours is the first predict-and-use model for sarcasm detection. We publicly release the code, gaze data, and our best models for further research. | [
"Tiwari, Divyank",
"Kanojia, Diptesh",
"Ray, Anupama",
"Nunna, Apoorva",
"Bhattacharyya, Pushpak"
] | Predict and Use: Harnessing Predicted Gaze to Improve Multimodal Sarcasm Detection | emnlp-main.988 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
||
https://aclanthology.org/2023.emnlp-main.989.bib | https://aclanthology.org/2023.emnlp-main.989/ | @inproceedings{wang-etal-2023-fine,
title = "Fine-grained Medical Vision-Language Representation Learning for Radiology Report Generation",
author = "Wang, Siyuan and
Peng, Bo and
Liu, Yichao and
Peng, Qi",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.989",
doi = "10.18653/v1/2023.emnlp-main.989",
pages = "15949--15956",
abstract = "Given the input radiology images, the objective of radiology report generation is to produce accurate and comprehensive medical reports, which typically include multiple descriptive clinical sentences associated with different phenotypes. Most existing works have relied on a pre-trained vision encoder to extract the visual representations of the images. In this study, we propose a phenotype-driven medical vision-language representation learning framework to efficiently bridge the gap between visual and textual modalities for improved text-oriented generation. In contrast to conventional methods which learn medical vision-language representations by contrasting images with entire reports, our approach learns more fine-grained representations by contrasting images with each sentence within the reports. The learned fine-grained representations can be used to improve radiology report generation. The experiments on two widely-used datasets MIMIC-CXR and IU X-ray demonstrate that our method can achieve promising performances and substantially outperform the conventional vision-language representation learning methods.",
}
| Given the input radiology images, the objective of radiology report generation is to produce accurate and comprehensive medical reports, which typically include multiple descriptive clinical sentences associated with different phenotypes. Most existing works have relied on a pre-trained vision encoder to extract the visual representations of the images. In this study, we propose a phenotype-driven medical vision-language representation learning framework to efficiently bridge the gap between visual and textual modalities for improved text-oriented generation. In contrast to conventional methods which learn medical vision-language representations by contrasting images with entire reports, our approach learns more fine-grained representations by contrasting images with each sentence within the reports. The learned fine-grained representations can be used to improve radiology report generation. The experiments on two widely-used datasets MIMIC-CXR and IU X-ray demonstrate that our method can achieve promising performances and substantially outperform the conventional vision-language representation learning methods. | [
"Wang, Siyuan",
"Peng, Bo",
"Liu, Yichao",
"Peng, Qi"
] | Fine-grained Medical Vision-Language Representation Learning for Radiology Report Generation | emnlp-main.989 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
||
https://aclanthology.org/2023.emnlp-main.990.bib | https://aclanthology.org/2023.emnlp-main.990/ | @inproceedings{liu-etal-2023-vit,
title = "{V}i{T}-{TTS}: Visual Text-to-Speech with Scalable Diffusion Transformer",
author = "Liu, Huadai and
Huang, Rongjie and
Lin, Xuan and
Xu, Wenqiang and
Zheng, Maozong and
Chen, Hong and
He, Jinzheng and
Zhao, Zhou",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.990",
doi = "10.18653/v1/2023.emnlp-main.990",
pages = "15957--15969",
abstract = "Text-to-speech(TTS) has undergone remarkable improvements in performance, particularly with the advent of Denoising Diffusion Probabilistic Models (DDPMs). However, the perceived quality of audio depends not solely on its content, pitch, rhythm, and energy, but also on the physical environment.In this work, we propose ViT-TTS, the first visual TTS model with scalable diffusion transformers. ViT-TTS complement the phoneme sequence with the visual information to generate high-perceived audio, opening up new avenues for practical applications of AR and VR to allow a more immersive and realistic audio experience. To mitigate the data scarcity in learning visual acoustic information, we 1) introduce a self-supervised learning framework to enhance both the visual-text encoder and denoiser decoder; 2) leverage the diffusion transformer scalable in terms of parameters and capacity to learn visual scene information. Experimental results demonstrate that ViT-TTS achieves new state-of-the-art results, outperforming cascaded systems and other baselines regardless of the visibility of the scene. With low-resource data (1h, 2h, 5h), ViT-TTS achieves comparative results with rich-resource baselines.",
}
| Text-to-speech(TTS) has undergone remarkable improvements in performance, particularly with the advent of Denoising Diffusion Probabilistic Models (DDPMs). However, the perceived quality of audio depends not solely on its content, pitch, rhythm, and energy, but also on the physical environment.In this work, we propose ViT-TTS, the first visual TTS model with scalable diffusion transformers. ViT-TTS complement the phoneme sequence with the visual information to generate high-perceived audio, opening up new avenues for practical applications of AR and VR to allow a more immersive and realistic audio experience. To mitigate the data scarcity in learning visual acoustic information, we 1) introduce a self-supervised learning framework to enhance both the visual-text encoder and denoiser decoder; 2) leverage the diffusion transformer scalable in terms of parameters and capacity to learn visual scene information. Experimental results demonstrate that ViT-TTS achieves new state-of-the-art results, outperforming cascaded systems and other baselines regardless of the visibility of the scene. With low-resource data (1h, 2h, 5h), ViT-TTS achieves comparative results with rich-resource baselines. | [
"Liu, Huadai",
"Huang, Rongjie",
"Lin, Xuan",
"Xu, Wenqiang",
"Zheng, Maozong",
"Chen, Hong",
"He, Jinzheng",
"Zhao, Zhou"
] | ViT-TTS: Visual Text-to-Speech with Scalable Diffusion Transformer | emnlp-main.990 | 2305.12708 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.991.bib | https://aclanthology.org/2023.emnlp-main.991/ | @inproceedings{jang-lukasiewicz-2023-consistency,
title = "Consistency Analysis of {C}hat{GPT}",
author = "Jang, Myeongjun and
Lukasiewicz, Thomas",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.991",
doi = "10.18653/v1/2023.emnlp-main.991",
pages = "15970--15985",
abstract = "ChatGPT has gained a huge popularity since its introduction. Its positive aspects have been reported through many media platforms, and some analyses even showed that ChatGPT achieved a decent grade in professional exams, adding extra support to the claim that AI can now assist and even replace humans in industrial fields. Others, however, doubt its reliability and trustworthiness. This paper investigates the trustworthiness of ChatGPT and GPT-4 regarding logically consistent behaviour, focusing specifically on semantic consistency and the properties of negation, symmetric, and transitive consistency. Our findings suggest that while both models appear to show an enhanced language understanding and reasoning ability, they still frequently fall short of generating logically consistent predictions. We also ascertain via experiments that prompt designing, few-shot learning and employing larger large language models (LLMs) are unlikely to be the ultimate solution to resolve the inconsistency issue of LLMs.",
}
| ChatGPT has gained a huge popularity since its introduction. Its positive aspects have been reported through many media platforms, and some analyses even showed that ChatGPT achieved a decent grade in professional exams, adding extra support to the claim that AI can now assist and even replace humans in industrial fields. Others, however, doubt its reliability and trustworthiness. This paper investigates the trustworthiness of ChatGPT and GPT-4 regarding logically consistent behaviour, focusing specifically on semantic consistency and the properties of negation, symmetric, and transitive consistency. Our findings suggest that while both models appear to show an enhanced language understanding and reasoning ability, they still frequently fall short of generating logically consistent predictions. We also ascertain via experiments that prompt designing, few-shot learning and employing larger large language models (LLMs) are unlikely to be the ultimate solution to resolve the inconsistency issue of LLMs. | [
"Jang, Myeongjun",
"Lukasiewicz, Thomas"
] | Consistency Analysis of ChatGPT | emnlp-main.991 | 2303.06273 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.992.bib | https://aclanthology.org/2023.emnlp-main.992/ | @inproceedings{van-der-meer-etal-2023-differences,
title = "Do Differences in Values Influence Disagreements in Online Discussions?",
author = "van der Meer, Michiel and
Vossen, Piek and
Jonker, Catholijn and
Murukannaiah, Pradeep",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.992",
doi = "10.18653/v1/2023.emnlp-main.992",
pages = "15986--16008",
abstract = "Disagreements are common in online discussions. Disagreement may foster collaboration and improve the quality of a discussion under some conditions. Although there exist methods for recognizing disagreement, a deeper understanding of factors that influence disagreement is lacking in the literature. We investigate a hypothesis that differences in \textit{personal values} are indicative of disagreement in online discussions. We show how state-of-the-art models can be used for estimating values in online discussions and how the estimated values can be aggregated into value profiles. We evaluate the estimated value profiles based on human-annotated agreement labels. We find that the dissimilarity of value profiles correlates with disagreement in specific cases. We also find that including value information in agreement prediction improves performance.",
}
| Disagreements are common in online discussions. Disagreement may foster collaboration and improve the quality of a discussion under some conditions. Although there exist methods for recognizing disagreement, a deeper understanding of factors that influence disagreement is lacking in the literature. We investigate a hypothesis that differences in \textit{personal values} are indicative of disagreement in online discussions. We show how state-of-the-art models can be used for estimating values in online discussions and how the estimated values can be aggregated into value profiles. We evaluate the estimated value profiles based on human-annotated agreement labels. We find that the dissimilarity of value profiles correlates with disagreement in specific cases. We also find that including value information in agreement prediction improves performance. | [
"van der Meer, Michiel",
"Vossen, Piek",
"Jonker, Catholijn",
"Murukannaiah, Pradeep"
] | Do Differences in Values Influence Disagreements in Online Discussions? | emnlp-main.992 | 2310.15757 | [
"https://github.com/m0re4u/value-disagreement"
] | https://huggingface.co/papers/2310.15757 | 1 | 0 | 0 | 4 | [] | [] | [] | 1 | Poster |
https://aclanthology.org/2023.emnlp-main.993.bib | https://aclanthology.org/2023.emnlp-main.993/ | @inproceedings{chamoun-etal-2023-automated,
title = "Automated Fact-Checking in Dialogue: Are Specialized Models Needed?",
author = "Chamoun, Eric and
Saeidi, Marzieh and
Vlachos, Andreas",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.993",
doi = "10.18653/v1/2023.emnlp-main.993",
pages = "16009--16020",
abstract = "Prior research has shown that typical fact-checking models for stand-alone claims struggle with claims made in conversation. As a solution, fine-tuning these models on dialogue data has been proposed. However, creating separate models for each use case is impractical, and we show that fine-tuning models for dialogue results in poor performance on typical fact-checking. To overcome this challenge, we present techniques that allow us to use the same models for both dialogue and typical fact-checking. These mainly focus on retrieval adaptation and transforming conversational inputs so that they can be accurately processed by models trained on stand-alone claims. We demonstrate that a typical fact-checking model incorporating these techniques is competitive with state-of-the-art models for dialogue, while maintaining its performance on stand-alone claims.",
}
| Prior research has shown that typical fact-checking models for stand-alone claims struggle with claims made in conversation. As a solution, fine-tuning these models on dialogue data has been proposed. However, creating separate models for each use case is impractical, and we show that fine-tuning models for dialogue results in poor performance on typical fact-checking. To overcome this challenge, we present techniques that allow us to use the same models for both dialogue and typical fact-checking. These mainly focus on retrieval adaptation and transforming conversational inputs so that they can be accurately processed by models trained on stand-alone claims. We demonstrate that a typical fact-checking model incorporating these techniques is competitive with state-of-the-art models for dialogue, while maintaining its performance on stand-alone claims. | [
"Chamoun, Eric",
"Saeidi, Marzieh",
"Vlachos, Andreas"
] | Automated Fact-Checking in Dialogue: Are Specialized Models Needed? | emnlp-main.993 | 2311.08195 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.994.bib | https://aclanthology.org/2023.emnlp-main.994/ | @inproceedings{gkoumas-etal-2023-digital,
title = "A Digital Language Coherence Marker for Monitoring Dementia",
author = "Gkoumas, Dimitris and
Tsakalidis, Adam and
Liakata, Maria",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.994",
doi = "10.18653/v1/2023.emnlp-main.994",
pages = "16021--16034",
abstract = "The use of spontaneous language to derive appropriate digital markers has become an emergent, promising and non-intrusive method to diagnose and monitor dementia. Here we propose methods to capture language coherence as a cost-effective, human-interpretable digital marker for monitoring cognitive changes in people with dementia. We introduce a novel task to learn the temporal logical consistency of utterances in short transcribed narratives and investigate a range of neural approaches. We compare such language coherence patterns between people with dementia and healthy controls and conduct a longitudinal evaluation against three clinical bio-markers to investigate the reliability of our proposed digital coherence marker. The coherence marker shows a significant difference between people with mild cognitive impairment, those with Alzheimer{'}s Disease and healthy controls. Moreover our analysis shows high association between the coherence marker and the clinical bio-markers as well as generalisability potential to other related conditions.",
}
| The use of spontaneous language to derive appropriate digital markers has become an emergent, promising and non-intrusive method to diagnose and monitor dementia. Here we propose methods to capture language coherence as a cost-effective, human-interpretable digital marker for monitoring cognitive changes in people with dementia. We introduce a novel task to learn the temporal logical consistency of utterances in short transcribed narratives and investigate a range of neural approaches. We compare such language coherence patterns between people with dementia and healthy controls and conduct a longitudinal evaluation against three clinical bio-markers to investigate the reliability of our proposed digital coherence marker. The coherence marker shows a significant difference between people with mild cognitive impairment, those with Alzheimer{'}s Disease and healthy controls. Moreover our analysis shows high association between the coherence marker and the clinical bio-markers as well as generalisability potential to other related conditions. | [
"Gkoumas, Dimitris",
"Tsakalidis, Adam",
"Liakata, Maria"
] | A Digital Language Coherence Marker for Monitoring Dementia | emnlp-main.994 | 2310.09623 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.995.bib | https://aclanthology.org/2023.emnlp-main.995/ | @inproceedings{wang-etal-2023-detecting,
title = "Detecting Spoilers in Movie Reviews with External Movie Knowledge and User Networks",
author = "Wang, Heng and
Zhang, Wenqian and
Bai, Yuyang and
Tan, Zhaoxuan and
Feng, Shangbin and
Zheng, Qinghua and
Luo, Minnan",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.995",
doi = "10.18653/v1/2023.emnlp-main.995",
pages = "16035--16050",
abstract = "Online movie review platforms are providing crowdsourced feedback for the film industry and the general public, while spoiler reviews greatly compromise user experience. Although preliminary research efforts were made to automatically identify spoilers, they merely focus on the review content itself, while robust spoiler detection requires putting the review into the context of facts and knowledge regarding movies, user behavior on film review platforms, and more. In light of these challenges, we first curate a large-scale network-based spoiler detection dataset LCS and a comprehensive and up-to-date movie knowledge base UKM. We then propose MVSD, a novel spoiler detection model that takes into account the external knowledge about movies and user activities on movie review platforms. Specifically, MVSD constructs three interconnecting heterogeneous information networks to model diverse data sources and their multi-view attributes, while we design and employ a novel heterogeneous graph neural network architecture for spoiler detection as node-level classification. Extensive experiments demonstrate that MVSD advances the state-of-the-art on two spoiler detection datasets, while the introduction of external knowledge and user interactions help ground robust spoiler detection.",
}
| Online movie review platforms are providing crowdsourced feedback for the film industry and the general public, while spoiler reviews greatly compromise user experience. Although preliminary research efforts were made to automatically identify spoilers, they merely focus on the review content itself, while robust spoiler detection requires putting the review into the context of facts and knowledge regarding movies, user behavior on film review platforms, and more. In light of these challenges, we first curate a large-scale network-based spoiler detection dataset LCS and a comprehensive and up-to-date movie knowledge base UKM. We then propose MVSD, a novel spoiler detection model that takes into account the external knowledge about movies and user activities on movie review platforms. Specifically, MVSD constructs three interconnecting heterogeneous information networks to model diverse data sources and their multi-view attributes, while we design and employ a novel heterogeneous graph neural network architecture for spoiler detection as node-level classification. Extensive experiments demonstrate that MVSD advances the state-of-the-art on two spoiler detection datasets, while the introduction of external knowledge and user interactions help ground robust spoiler detection. | [
"Wang, Heng",
"Zhang, Wenqian",
"Bai, Yuyang",
"Tan, Zhaoxuan",
"Feng, Shangbin",
"Zheng, Qinghua",
"Luo, Minnan"
] | Detecting Spoilers in Movie Reviews with External Movie Knowledge and User Networks | emnlp-main.995 | 2304.11411 | [
"https://github.com/arthur-heng/spoiler-detection"
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.996.bib | https://aclanthology.org/2023.emnlp-main.996/ | @inproceedings{li-etal-2023-joyful,
title = "Joyful: Joint Modality Fusion and Graph Contrastive Learning for Multimoda Emotion Recognition",
author = "Li, Dongyuan and
Wang, Yusong and
Funakoshi, Kotaro and
Okumura, Manabu",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.996",
doi = "10.18653/v1/2023.emnlp-main.996",
pages = "16051--16069",
abstract = "Multimodal emotion recognition aims to recognize emotions for each utterance from multiple modalities, which has received increasing attention for its application in human-machine interaction. Current graph-based methods fail to simultaneously depict global contextual features and local diverse uni-modal features in a dialogue. Furthermore, with the number of graph layers increasing, they easily fall into over-smoothing. In this paper, we propose a method for joint modality fusion and graph contrastive learning for multimodal emotion recognition (Joyful), where multimodality fusion, contrastive learning, and emotion recognition are jointly optimized. Specifically, we first design a new multimodal fusion mechanism that can provide deep interaction and fusion between the global contextual and uni-modal specific features. Then, we introduce a graph contrastive learning framework with inter- and intra-view contrastive losses to learn more distinguishable representations for samples with different sentiments. Extensive experiments on three benchmark datasets indicate that Joyful achieved state-of-the-art (SOTA) performance compared with all baselines. Code is released on Github (https://anonymous.4open.science/r/MERC-7F88).",
}
| Multimodal emotion recognition aims to recognize emotions for each utterance from multiple modalities, which has received increasing attention for its application in human-machine interaction. Current graph-based methods fail to simultaneously depict global contextual features and local diverse uni-modal features in a dialogue. Furthermore, with the number of graph layers increasing, they easily fall into over-smoothing. In this paper, we propose a method for joint modality fusion and graph contrastive learning for multimodal emotion recognition (Joyful), where multimodality fusion, contrastive learning, and emotion recognition are jointly optimized. Specifically, we first design a new multimodal fusion mechanism that can provide deep interaction and fusion between the global contextual and uni-modal specific features. Then, we introduce a graph contrastive learning framework with inter- and intra-view contrastive losses to learn more distinguishable representations for samples with different sentiments. Extensive experiments on three benchmark datasets indicate that Joyful achieved state-of-the-art (SOTA) performance compared with all baselines. Code is released on Github (https://anonymous.4open.science/r/MERC-7F88). | [
"Li, Dongyuan",
"Wang, Yusong",
"Funakoshi, Kotaro",
"Okumura, Manabu"
] | Joyful: Joint Modality Fusion and Graph Contrastive Learning for Multimoda Emotion Recognition | emnlp-main.996 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
||
https://aclanthology.org/2023.emnlp-main.997.bib | https://aclanthology.org/2023.emnlp-main.997/ | @inproceedings{song-etal-2023-hyperrank,
title = "{H}yper{R}ank: Hyperbolic Ranking Model for Unsupervised Keyphrase Extraction",
author = "Song, Mingyang and
Liu, Huafeng and
Jing, Liping",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.997",
doi = "10.18653/v1/2023.emnlp-main.997",
pages = "16070--16080",
abstract = "Given the exponential growth in the number of documents on the web in recent years, there is an increasing demand for accurate models to extract keyphrases from such documents. Keyphrase extraction is the task of automatically identifying representative keyphrases from the source document. Typically, candidate keyphrases exhibit latent hierarchical structures embedded with intricate syntactic and semantic information. Moreover, the relationships between candidate keyphrases and the document also form hierarchical structures. Therefore, it is essential to consider these latent hierarchical structures when extracting keyphrases. However, many recent unsupervised keyphrase extraction models overlook this aspect, resulting in incorrect keyphrase extraction. In this paper, we address this issue by proposing a new hyperbolic ranking model (HyperRank). HyperRank is designed to jointly model global and local context information for estimating the importance of each candidate keyphrase within the hyperbolic space, enabling accurate keyphrase extraction. Experimental results demonstrate that HyperRank significantly outperforms recent state-of-the-art baselines.",
}
| Given the exponential growth in the number of documents on the web in recent years, there is an increasing demand for accurate models to extract keyphrases from such documents. Keyphrase extraction is the task of automatically identifying representative keyphrases from the source document. Typically, candidate keyphrases exhibit latent hierarchical structures embedded with intricate syntactic and semantic information. Moreover, the relationships between candidate keyphrases and the document also form hierarchical structures. Therefore, it is essential to consider these latent hierarchical structures when extracting keyphrases. However, many recent unsupervised keyphrase extraction models overlook this aspect, resulting in incorrect keyphrase extraction. In this paper, we address this issue by proposing a new hyperbolic ranking model (HyperRank). HyperRank is designed to jointly model global and local context information for estimating the importance of each candidate keyphrase within the hyperbolic space, enabling accurate keyphrase extraction. Experimental results demonstrate that HyperRank significantly outperforms recent state-of-the-art baselines. | [
"Song, Mingyang",
"Liu, Huafeng",
"Jing, Liping"
] | HyperRank: Hyperbolic Ranking Model for Unsupervised Keyphrase Extraction | emnlp-main.997 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
||
https://aclanthology.org/2023.emnlp-main.998.bib | https://aclanthology.org/2023.emnlp-main.998/ | @inproceedings{zacharopoulos-etal-2023-assessing,
title = "Assessing the influence of attractor-verb distance on grammatical agreement in humans and language models",
author = "Zacharopoulos, Christos and
Desbordes, Th{\'e}o and
Sabl{\'e}-Meyer, Mathias",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.998",
doi = "10.18653/v1/2023.emnlp-main.998",
pages = "16081--16090",
abstract = "Subject-verb agreement in the presence of an attractor noun located between the main noun and the verb elicits complex behavior: judgments of grammaticality are modulated by the grammatical features of the attractor. For example, in the sentence $\textit{``The girl near the boys likes climbing''}$, the attractor ($\textit{boys}$) disagrees in grammatical number with the verb ($\textit{likes}$), creating a locally implausible transition probability. Here, we parametrically modulate the distance between the attractor and the verb while keeping the length of the sentence equal. We evaluate the performance of both humans and two artificial neural network models: both make more mistakes when the attractor is closer to the verb, but neural networks get close to the chance level while humans are mostly able to overcome the attractor interference. Additionally, we report a linear effect of attractor distance on reaction times. We hypothesize that a possible reason for the proximity effect is the calculation of transition probabilities between adjacent words. Nevertheless, classical models of attraction such as the cue-based model might suffice to explain this phenomenon, thus paving the way for new research. Data and analyses available at https://osf.io/d4g6k",
}
| Subject-verb agreement in the presence of an attractor noun located between the main noun and the verb elicits complex behavior: judgments of grammaticality are modulated by the grammatical features of the attractor. For example, in the sentence $\textit{``The girl near the boys likes climbing''}$, the attractor ($\textit{boys}$) disagrees in grammatical number with the verb ($\textit{likes}$), creating a locally implausible transition probability. Here, we parametrically modulate the distance between the attractor and the verb while keeping the length of the sentence equal. We evaluate the performance of both humans and two artificial neural network models: both make more mistakes when the attractor is closer to the verb, but neural networks get close to the chance level while humans are mostly able to overcome the attractor interference. Additionally, we report a linear effect of attractor distance on reaction times. We hypothesize that a possible reason for the proximity effect is the calculation of transition probabilities between adjacent words. Nevertheless, classical models of attraction such as the cue-based model might suffice to explain this phenomenon, thus paving the way for new research. Data and analyses available at https://osf.io/d4g6k | [
"Zacharopoulos, Christos",
"Desbordes, Th{\\'e}o",
"Sabl{\\'e}-Meyer, Mathias"
] | Assessing the influence of attractor-verb distance on grammatical agreement in humans and language models | emnlp-main.998 | 2311.16978 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
|
https://aclanthology.org/2023.emnlp-main.999.bib | https://aclanthology.org/2023.emnlp-main.999/ | @inproceedings{singh-etal-2023-federated,
title = "Federated Meta-Learning for Emotion and Sentiment Aware Multi-modal Complaint Identification",
author = "Singh, Apoorva and
Chandrasekar, Siddarth and
Saha, Sriparna and
Sen, Tanmay",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.999",
doi = "10.18653/v1/2023.emnlp-main.999",
pages = "16091--16103",
abstract = "Automatic detection of consumers{'} complaints about items or services they buy can be critical for organizations and online merchants. Previous studies on complaint identification are limited to text. Images along with the reviews can provide cues to identify complaints better, thus emphasizing the importance of incorporating multi-modal inputs into the process. Generally, the customer{'}s emotional state significantly impacts the complaint expression; thus, the effect of emotion and sentiment on complaint identification must also be investigated. Furthermore, different organizations are usually not allowed to share their privacy-sensitive records due to data security and privacy concerns. Due to these issues, traditional models find it hard to understand and identify complaint patterns, particularly in the financial and healthcare sectors. In this work, we created a new dataset - Multi-modal Complaint Dataset (MCD), a collection of reviews and images of the products posted on the website of the retail giant Amazon. We propose a federated meta-learning-based multi-modal multi-task framework for identifying complaints considering emotion recognition and sentiment analysis as two auxiliary tasks. Experimental results indicate that the proposed approach outperforms the baselines and the state-of-the-art approaches in centralized and federated meta-learning settings.",
}
| Automatic detection of consumers{'} complaints about items or services they buy can be critical for organizations and online merchants. Previous studies on complaint identification are limited to text. Images along with the reviews can provide cues to identify complaints better, thus emphasizing the importance of incorporating multi-modal inputs into the process. Generally, the customer{'}s emotional state significantly impacts the complaint expression; thus, the effect of emotion and sentiment on complaint identification must also be investigated. Furthermore, different organizations are usually not allowed to share their privacy-sensitive records due to data security and privacy concerns. Due to these issues, traditional models find it hard to understand and identify complaint patterns, particularly in the financial and healthcare sectors. In this work, we created a new dataset - Multi-modal Complaint Dataset (MCD), a collection of reviews and images of the products posted on the website of the retail giant Amazon. We propose a federated meta-learning-based multi-modal multi-task framework for identifying complaints considering emotion recognition and sentiment analysis as two auxiliary tasks. Experimental results indicate that the proposed approach outperforms the baselines and the state-of-the-art approaches in centralized and federated meta-learning settings. | [
"Singh, Apoorva",
"Ch",
"rasekar, Siddarth",
"Saha, Sriparna",
"Sen, Tanmay"
] | Federated Meta-Learning for Emotion and Sentiment Aware Multi-modal Complaint Identification | emnlp-main.999 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
||
https://aclanthology.org/2023.emnlp-main.1000.bib | https://aclanthology.org/2023.emnlp-main.1000/ | @inproceedings{perez-etal-2023-semantic,
title = "Semantic Similarity Models for Depression Severity Estimation",
author = "P{\'e}rez, Anxo and
Warikoo, Neha and
Wang, Kexin and
Parapar, Javier and
Gurevych, Iryna",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.1000",
doi = "10.18653/v1/2023.emnlp-main.1000",
pages = "16104--16118",
abstract = "Depressive disorders constitute a severe public health issue worldwide. However, public health systems have limited capacity for case detection and diagnosis. In this regard, the widespread use of social media has opened up a way to access public information on a large scale. Computational methods can serve as support tools for rapid screening by exploiting this user-generated social media content. This paper presents an efficient semantic pipeline to study depression severity in individuals based on their social media writings. We select test user sentences for producing semantic rankings over an index of representative training sentences corresponding to depressive symptoms and severity levels. Then, we use the sentences from those results as evidence for predicting symptoms severity. For that, we explore different aggregation methods to answer one of four Beck Depression Inventory (BDI-II) options per symptom. We evaluate our methods on two Reddit-based benchmarks, achieving improvement over state of the art in terms of measuring depression level.",
}
| Depressive disorders constitute a severe public health issue worldwide. However, public health systems have limited capacity for case detection and diagnosis. In this regard, the widespread use of social media has opened up a way to access public information on a large scale. Computational methods can serve as support tools for rapid screening by exploiting this user-generated social media content. This paper presents an efficient semantic pipeline to study depression severity in individuals based on their social media writings. We select test user sentences for producing semantic rankings over an index of representative training sentences corresponding to depressive symptoms and severity levels. Then, we use the sentences from those results as evidence for predicting symptoms severity. For that, we explore different aggregation methods to answer one of four Beck Depression Inventory (BDI-II) options per symptom. We evaluate our methods on two Reddit-based benchmarks, achieving improvement over state of the art in terms of measuring depression level. | [
"P{\\'e}rez, Anxo",
"Warikoo, Neha",
"Wang, Kexin",
"Parapar, Javier",
"Gurevych, Iryna"
] | Semantic Similarity Models for Depression Severity Estimation | emnlp-main.1000 | 2211.07624 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | 0 | Poster |
Subsets and Splits