CivilSum: A Dataset for Abstractive Summarization of Indian Court Decisions

Anonymous

Abstract: We introduce \textsc{CivilSum}, a dataset of 23,350 legal case decisions paired with human-written abstractive summaries from the Supreme Court of India and Indian High Courts. In contrast to other domains such as news articles, our analysis shows the most important content tends to appear at the end of the documents. We measure the effect of this \emph{tail bias} on summarization performance using strong baselines for long-document abstractive summarization, and the results highlight the importance of long sequence modeling for the proposed task. \textsc{CivilSum} and related code are publicly available for research purposes.

Paper Type: short

Research Area: Resources and Evaluation

Contribution Types: Data resources

Languages Studied: English

Consent To Share Submission Details: On behalf of all authors, we agree to the terms above to share our submission details.