A novel dataset for benchmarking citation worthiness detection task in the American Legal Corpus. For more details about the dataset please refer to the original paper.
Data Fields
- File Name: the case file to which the sentence belongs.
- Sentence Number: The sentence number as present in the document.
- Sentence: The naturally occurring sentence in the text (after preprocessing/removing citation span.)
- Label: Integer value of ‘0’ or ‘1’. ‘0’ represents that the sentence is not citation worthy whereas ‘1’ represents that the sentence is citation worthy.
Data Splits
| Split | #datapoints |
|---|---|
| Train-Small | 800,000 |
| Validation-Small | 100,000 |
| Test-Small | 100,000 |
| Train-Medium | 8,000,000 |
| Validation-Medium | 1,000,000 |
| Test-Medium | 1,000,000 |
| Train-Large | 142,588,927 |
| Validation-Large | 17,934,940 |
| Test-Large | 17,935,336 |
Small Dataset
from datasets import load_dataset
# get small dataset
dataset = load_dataset("Vidhaan/LegalCitationWorthiness", "small")
Medium Dataset
from datasets import load_dataset
# get medium dataset
dataset = load_dataset("Vidhaan/LegalCitationWorthiness", "medium")
Large Dataset
from datasets import load_dataset
# get large dataset
dataset = load_dataset("Vidhaan/LegalCitationWorthiness", "large")
Citation Information
Contributions
Thanks to @PritishWadhwa, @gitongithub, @khatrimann, @reshma, @dhumketu for adding this dataset