Legal Extractive Summarization of U.S. Court Opinions

TL;DR

This paper presents a reinforcement learning-based extractive summarization model for U.S. court opinions, demonstrating superior automated and human-evaluated performance and open-sourcing the models to improve legal accessibility.

Contribution

Introduces MemSum, a reinforcement learning model for legal extractive summarization that outperforms transformer models and is publicly available.

Findings

MemSum outperforms transformer-based models in automated metrics.

Human evaluation confirms MemSum effectively captures key points.

Open-sourcing promotes legal accessibility and democratization.

Abstract

This paper tackles the task of legal extractive summarization using a dataset of 430K U.S. court opinions with key passages annotated. According to automated summary quality metrics, the reinforcement-learning-based MemSum model is best and even out-performs transformer-based models. In turn, expert human evaluation shows that MemSum summaries effectively capture the key points of lengthy court opinions. Motivated by these results, we open-source our models to the general public. This represents progress towards democratizing law and making U.S. court opinions more accessible to the general public.