Update README.md
Browse files
README.md
CHANGED
@@ -9,6 +9,11 @@ language:
|
|
9 |
- ar
|
10 |
- hi
|
11 |
- es
|
|
|
|
|
|
|
|
|
|
|
12 |
license: openrail++
|
13 |
size_categories:
|
14 |
- 1K<n<10K
|
@@ -72,7 +77,7 @@ configs:
|
|
72 |
- split: hi
|
73 |
path: data/hi-*
|
74 |
---
|
75 |
-
**
|
76 |
|
77 |
[](https://aclanthology.org/2025.coling-main.535/)
|
78 |
[](https://ceur-ws.org/Vol-3740/paper-223.pdf)
|
@@ -82,11 +87,15 @@ For each of 9 languages, we collected 1k pairs of toxic<->detoxified instances s
|
|
82 |
|
83 |
📰 **Updates**
|
84 |
|
|
|
|
|
|
|
|
|
85 |
**[2025]** We dived into the explainability of our data in our new [COLING paper](https://huggingface.co/papers/2412.11691)!
|
86 |
|
87 |
**[2024]** You can check additional releases for [Ukrainian ParaDetox](https://huggingface.co/datasets/textdetox/uk_paradetox) and [Spanish ParaDetox](https://huggingface.co/datasets/textdetox/es_paradetox) from NAACL 2024!
|
88 |
|
89 |
-
**[2024]** **April, 23rd, update: We are realsing the parallel
|
90 |
|
91 |
**[2022]** You can also check previously created training corpora: [English ParaDetox](https://huggingface.co/datasets/s-nlp/paradetox) from ACL 2022 and [Russian ParaDetox](https://huggingface.co/datasets/s-nlp/ru_paradetox).
|
92 |
|
@@ -101,6 +110,12 @@ The list of the sources for the original toxic sentences:
|
|
101 |
* Amhairc: [Amharic Hate Speech](https://github.com/uhh-lt/AmharicHateSpeech)
|
102 |
* Arabic: [OSACT4](https://edinburghnlp.inf.ed.ac.uk/workshops/OSACT4/)
|
103 |
* Hindi: [Hostility Detection Dataset in Hindi](https://competitions.codalab.org/competitions/26654#learn_the_details-dataset), [Overview of the HASOC track at FIRE 2019: Hate Speech and Offensive Content Identification in Indo-European Languages](https://dl.acm.org/doi/pdf/10.1145/3368567.3368584?download=true)
|
|
|
|
|
|
|
|
|
|
|
|
|
104 |
|
105 |
## Citation
|
106 |
If you would like to acknowledge our work, please, cite the following manuscripts:
|
|
|
9 |
- ar
|
10 |
- hi
|
11 |
- es
|
12 |
+
- it
|
13 |
+
- fr
|
14 |
+
- he
|
15 |
+
- ja
|
16 |
+
- tt
|
17 |
license: openrail++
|
18 |
size_categories:
|
19 |
- 1K<n<10K
|
|
|
77 |
- split: hi
|
78 |
path: data/hi-*
|
79 |
---
|
80 |
+
**Multilingual Text Detoxification with Parallel Data**
|
81 |
|
82 |
[](https://aclanthology.org/2025.coling-main.535/)
|
83 |
[](https://ceur-ws.org/Vol-3740/paper-223.pdf)
|
|
|
87 |
|
88 |
📰 **Updates**
|
89 |
|
90 |
+
**[2025]** The second edition of TextDetox shared task! [webpage](https://pan.webis.de/clef25/pan25-web/text-detoxification.html)
|
91 |
+
|
92 |
+
**[2025]** We extend our data to new languages! Now also included: Italian, French, Hebrew, Hinglish, Japanese, Tatar. Check our [test](https://huggingface.co/datasets/textdetox/multilingual_paradetox_test) part.
|
93 |
+
|
94 |
**[2025]** We dived into the explainability of our data in our new [COLING paper](https://huggingface.co/papers/2412.11691)!
|
95 |
|
96 |
**[2024]** You can check additional releases for [Ukrainian ParaDetox](https://huggingface.co/datasets/textdetox/uk_paradetox) and [Spanish ParaDetox](https://huggingface.co/datasets/textdetox/es_paradetox) from NAACL 2024!
|
97 |
|
98 |
+
**[2024]** **April, 23rd, update: We are realsing the parallel train set! The test part for the final phase of the competition is available [here](https://huggingface.co/datasets/textdetox/multilingual_paradetox_test)!!!**
|
99 |
|
100 |
**[2022]** You can also check previously created training corpora: [English ParaDetox](https://huggingface.co/datasets/s-nlp/paradetox) from ACL 2022 and [Russian ParaDetox](https://huggingface.co/datasets/s-nlp/ru_paradetox).
|
101 |
|
|
|
110 |
* Amhairc: [Amharic Hate Speech](https://github.com/uhh-lt/AmharicHateSpeech)
|
111 |
* Arabic: [OSACT4](https://edinburghnlp.inf.ed.ac.uk/workshops/OSACT4/)
|
112 |
* Hindi: [Hostility Detection Dataset in Hindi](https://competitions.codalab.org/competitions/26654#learn_the_details-dataset), [Overview of the HASOC track at FIRE 2019: Hate Speech and Offensive Content Identification in Indo-European Languages](https://dl.acm.org/doi/pdf/10.1145/3368567.3368584?download=true)
|
113 |
+
* Italian: [AMI](https://github.com/dnozza/ami2020), [HODI](https://github.com/HODI-EVALITA/HODI_2023), [Jigsaw Multilingual Toxic Comment](https://www.kaggle.com/competitions/jigsaw-multilingual-toxic-comment-classification/overview)
|
114 |
+
* French: [FrenchToxicityPrompts](https://europe.naverlabs.com/research/publications/frenchtoxicityprompts-a-large-benchmark-for-evaluating-and-mitigating-toxicity-in-french-texts/), [Jigsaw Multilingual Toxic Comment](https://www.kaggle.com/competitions/jigsaw-multilingual-toxic-comment-classification/overview)
|
115 |
+
* Hebrew: [Hebrew Offensive Language Dataset](https://github.com/NataliaVanetik/HebrewOffensiveLanguageDatasetForTheDetoxificationProject/tree/main)
|
116 |
+
* Hinglish: [Hinglish Hate Detection](https://github.com/victor7246/Hinglish_Hate_Detection/blob/main/data/raw/trac1-dataset/hindi/agr_hi_dev.csv)
|
117 |
+
* Japanese: posts from [2chan](https://huggingface.co/datasets/p1atdev/open2ch)
|
118 |
+
* Tatar: ours.
|
119 |
|
120 |
## Citation
|
121 |
If you would like to acknowledge our work, please, cite the following manuscripts:
|