sihao commited on
Commit
130bb64
·
1 Parent(s): df7e00b

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +22 -68
README.md CHANGED
@@ -27,85 +27,39 @@ size_categories:
27
  ### Dataset Summary
28
 
29
  This is a reproduced (i.e. after web-crawling) and processed version of [the "PropSegment" dataset](https://github.com/google-research-datasets/PropSegmEnt) from Google Research.
 
 
 
 
30
  PropSegment (Proposition-level Segmentation and Entailment) is a large-scale, human annotated dataset for segmenting English text into propositions, and recognizing proposition-level entailment relations --- whether a different, related document entails each proposition, contradicts it, or neither.
31
 
32
- The dataset features >45k human annotated propositions, i.e. individual semantic units within sentences, as well as >35k entailment labels between propositions and documents.
33
 
34
  Check out more details in the [dataset paper](https://arxiv.org/abs/2212.10750).
35
 
36
 
37
  ## Dataset Structure
38
 
39
- ### Data Instances
40
-
41
-
42
-
43
- ### Data Fields
44
-
45
- [More Information Needed]
46
-
47
- ### Data Splits
48
-
49
- [More Information Needed]
50
-
51
- ## Dataset Creation
52
-
53
- ### Curation Rationale
54
-
55
- [More Information Needed]
56
-
57
- ### Source Data
58
-
59
- #### Initial Data Collection and Normalization
60
-
61
- [More Information Needed]
62
-
63
- #### Who are the source language producers?
64
-
65
- [More Information Needed]
66
-
67
- ### Annotations
68
-
69
- #### Annotation process
70
-
71
- [More Information Needed]
72
-
73
- #### Who are the annotators?
74
-
75
- [More Information Needed]
76
-
77
- ### Personal and Sensitive Information
78
-
79
- [More Information Needed]
80
-
81
- ## Considerations for Using the Data
82
-
83
- ### Social Impact of Dataset
84
-
85
- [More Information Needed]
86
-
87
- ### Discussion of Biases
88
-
89
- [More Information Needed]
90
-
91
- ### Other Known Limitations
92
-
93
- [More Information Needed]
94
-
95
- ## Additional Information
96
-
97
- ### Dataset Curators
98
-
99
- [More Information Needed]
100
 
101
- ### Licensing Information
 
 
102
 
103
- [More Information Needed]
 
 
 
 
 
104
 
105
- ### Citation Information
106
 
107
- [More Information Needed]
108
 
109
- ### Contributions
 
 
 
 
 
110
 
111
- [More Information Needed]
 
27
  ### Dataset Summary
28
 
29
  This is a reproduced (i.e. after web-crawling) and processed version of [the "PropSegment" dataset](https://github.com/google-research-datasets/PropSegmEnt) from Google Research.
30
+
31
+ Since the [`News`](https://github.com/google-research-datasets/NewSHead) portion of the dataset is released only via urls, we reconstruct the dataset by crawling.
32
+ Overall, ~96% of the dataset can be reproduced, and the rest ~4% either have url no longer valid, or sentences that have been edited (i.e. cannot be aligned with the orignial dataset).
33
+
34
  PropSegment (Proposition-level Segmentation and Entailment) is a large-scale, human annotated dataset for segmenting English text into propositions, and recognizing proposition-level entailment relations --- whether a different, related document entails each proposition, contradicts it, or neither.
35
 
36
+ The original dataset features >45k human annotated propositions, i.e. individual semantic units within sentences, as well as >35k entailment labels between propositions and documents.
37
 
38
  Check out more details in the [dataset paper](https://arxiv.org/abs/2212.10750).
39
 
40
 
41
  ## Dataset Structure
42
 
43
+ Here we provide processed versions of the dataset for seq2seq model inputs/outputs.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
44
 
45
+ `proposition_segmentation.*.jsonl` contains data for the text segmentation task, i.e. split a sentence into propositions.
46
+ The output propositions are concatenated as one string (with no particular order between them) by a special token `[SEP]`.
47
+ Each proposition is annotated as spans enclosed by `[M]` and `[/M]`.
48
 
49
+ ```
50
+ {
51
+ "sentence": "This film marks the directorial debut for production designer Robert Stromberg.",
52
+ "propositions": "This film marks the directorial debut for [M]production designer Robert Stromberg.[/M][SEP]This [M]film marks the directorial debut for[/M] production designer [M]Robert Stromberg[/M]."
53
+ }
54
+ ```
55
 
 
56
 
57
+ ### Citation
58
 
59
+ @inproceedings{chen2023propsegment,
60
+ title = "{PropSegmEnt}: A Large-Scale Corpus for Proposition-Level Segmentation and Entailment Recognition",
61
+ author = "Chen, Sihao and Buthpitiya, Senaka and Fabrikant, Alex and Roth, Dan and Schuster, Tal",
62
+ booktitle = "Findings of the Association for Computational Linguistics: ACL 2023",
63
+ year = "2023",
64
+ }
65