Commit
·
0be35e2
1
Parent(s):
75ff642
Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,50 @@
|
|
1 |
---
|
2 |
license: gpl-3.0
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
license: gpl-3.0
|
3 |
---
|
4 |
+
## Dataset Description
|
5 |
+
|
6 |
+
- **Homepage:** https://www.darrow.ai/
|
7 |
+
- **Repository:** https://github.com/darrow-labs/ClassActionPrediction
|
8 |
+
- **Paper:** https://arxiv.org/abs/2110.
|
9 |
+
- **Leaderboard:** N/A
|
10 |
+
- **Point of Contact:** [Gila Hayat](mailto:[email protected])
|
11 |
+
|
12 |
+
### Dataset Summary
|
13 |
+
|
14 |
+
USClassActions is an English dataset of 200 complaints from the US Federal Court with the respective binarized judgment outcome (Win/Lose). The dataset poses a challenging text classification task. We are happy to share this dataset in order to promote robustness and fairness studies on the critical area of legal NLP.
|
15 |
+
|
16 |
+
### Data Instances
|
17 |
+
```python
|
18 |
+
from datasets import load_dataset
|
19 |
+
dataset = load_dataset('darrow-ai/USClassActionOutcomes_ExpertsAnnotations')
|
20 |
+
```
|
21 |
+
### Data Fields
|
22 |
+
`id`: (**int**) a unique identifier of the document \
|
23 |
+
`origin_label `: (**str**) the outcome of the case \
|
24 |
+
`target_text`: (**str**) the facts of the case \
|
25 |
+
`annotator_prediction `: (**str**) annotators predictions of the case outcome based on the target_text \
|
26 |
+
`annotator_confidence `: (**str**) the annotator's level of confidence in his outcome prediction \
|
27 |
+
|
28 |
+
### Curation Rationale
|
29 |
+
|
30 |
+
The dataset was curated by Darrow.ai (2022).
|
31 |
+
|
32 |
+
### Citation Information
|
33 |
+
|
34 |
+
*Gil Semo, Dor Bernsohn, Ben Hagag, Gila Hayat, and Joel Niklaus*
|
35 |
+
*ClassActionPrediction: A Challenging Benchmark for Legal Judgment Prediction of Class Action Cases in the US*
|
36 |
+
*Proceedings of the 2022 Natural Legal Language Processing Workshop. Abu Dhabi. 2022*
|
37 |
+
```
|
38 |
+
@InProceedings{darrow-niklaus-2022-uscp,
|
39 |
+
author = {Semo, Gil
|
40 |
+
and Bernsohn, Dor
|
41 |
+
and Stürmer, Matthias
|
42 |
+
and Hagag, Ben
|
43 |
+
and Hayat, Gila
|
44 |
+
and Niklaus, Joel},
|
45 |
+
title = {ClassActionPrediction: A Challenging Benchmark for Legal Judgment Prediction of Class Action Cases in the US},
|
46 |
+
booktitle = {Proceedings of the 2022 Natural Legal Language Processing Workshop},
|
47 |
+
year = {2022},
|
48 |
+
location = {Abu Dhabi},
|
49 |
+
}
|
50 |
+
```
|