Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Low-Resource Court Judgment Summarization for Common Law Systems

Shuaiqi Liu, Jiannong Cao, Yicong Li, Ruosong Yang, Zhiyuan Wen

TL;DR

The paper addresses cross-jurisdiction court judgment summarization in resource-constrained settings by introducing CLSum, a four-jurisdiction dataset (Canada, Australia, United Kingdom, Hong Kong) and a two-stage, LLM-assisted summarization framework designed for long, complex judgments. It proposes data augmentation with legal-knowledge constraints and a LTScore evaluation metric to better capture legal terminology usage, alongside memory-efficient training and content-selection strategies. Empirical results show that LLM-based methods can perform well in few-shot and zero-shot scenarios, while data augmentation and model design choices (content selection, adapters, RLHF) significantly impact performance, particularly in data-scarce subsets. The work advances practical cross-jurisdiction judgment summarization and lays groundwork for scalable, legally aware automatic summarization in common-law systems, with implications for practitioners and public accessibility to judgments.

Abstract

Common law courts need to refer to similar precedents' judgments to inform their current decisions. Generating high-quality summaries of court judgment documents can facilitate legal practitioners to efficiently review previous cases and assist the general public in accessing how the courts operate and how the law is applied. Previous court judgment summarization research focuses on civil law or a particular jurisdiction's judgments. However, judges can refer to the judgments from all common law jurisdictions. Current summarization datasets are insufficient to satisfy the demands of summarizing precedents across multiple jurisdictions, especially when labeled data are scarce for many jurisdictions. To address the lack of datasets, we present CLSum, the first dataset for summarizing multi-jurisdictional common law court judgment documents. Besides, this is the first court judgment summarization work adopting large language models (LLMs) in data augmentation, summary generation, and evaluation. Specifically, we design an LLM-based data augmentation method incorporating legal knowledge. We also propose a legal knowledge enhanced evaluation metric based on LLM to assess the quality of generated judgment summaries. Our experimental results verify that the LLM-based summarization methods can perform well in the few-shot and zero-shot settings. Our LLM-based data augmentation method can mitigate the impact of low data resources. Furthermore, we carry out comprehensive comparative experiments to find essential model components and settings that are capable of enhancing summarization performance.

Low-Resource Court Judgment Summarization for Common Law Systems

TL;DR

The paper addresses cross-jurisdiction court judgment summarization in resource-constrained settings by introducing CLSum, a four-jurisdiction dataset (Canada, Australia, United Kingdom, Hong Kong) and a two-stage, LLM-assisted summarization framework designed for long, complex judgments. It proposes data augmentation with legal-knowledge constraints and a LTScore evaluation metric to better capture legal terminology usage, alongside memory-efficient training and content-selection strategies. Empirical results show that LLM-based methods can perform well in few-shot and zero-shot scenarios, while data augmentation and model design choices (content selection, adapters, RLHF) significantly impact performance, particularly in data-scarce subsets. The work advances practical cross-jurisdiction judgment summarization and lays groundwork for scalable, legally aware automatic summarization in common-law systems, with implications for practitioners and public accessibility to judgments.

Abstract

Common law courts need to refer to similar precedents' judgments to inform their current decisions. Generating high-quality summaries of court judgment documents can facilitate legal practitioners to efficiently review previous cases and assist the general public in accessing how the courts operate and how the law is applied. Previous court judgment summarization research focuses on civil law or a particular jurisdiction's judgments. However, judges can refer to the judgments from all common law jurisdictions. Current summarization datasets are insufficient to satisfy the demands of summarizing precedents across multiple jurisdictions, especially when labeled data are scarce for many jurisdictions. To address the lack of datasets, we present CLSum, the first dataset for summarizing multi-jurisdictional common law court judgment documents. Besides, this is the first court judgment summarization work adopting large language models (LLMs) in data augmentation, summary generation, and evaluation. Specifically, we design an LLM-based data augmentation method incorporating legal knowledge. We also propose a legal knowledge enhanced evaluation metric based on LLM to assess the quality of generated judgment summaries. Our experimental results verify that the LLM-based summarization methods can perform well in the few-shot and zero-shot settings. Our LLM-based data augmentation method can mitigate the impact of low data resources. Furthermore, we carry out comprehensive comparative experiments to find essential model components and settings that are capable of enhancing summarization performance.
Paper Structure (25 sections, 3 equations, 8 figures, 10 tables)

This paper contains 25 sections, 3 equations, 8 figures, 10 tables.

Table of Contents

  1. Introduction
  2. Objectives and Contribution
  3. Related Work
  4. Court Judgment Summarization Datasets
  5. Court Judgment Summarization Methods
  6. CLSum Dataset
  7. Collecting and Pre-processing Data
  8. Description of the CLSum's Subsets
  9. Dataset Analysis
  10. Method
  11. Mitigating the Impact of Insufficient Labeled Samples
  12. Salient Content Identification and Integration
  13. Improving the Efficiency of Models and Training Methods
  14. Experiments
  15. Baselines
  16. ...and 10 more sections

Figures (8)

  • Figure 1: Our workflow of Court Judgment Summarization.
  • Figure 2: Distributions of extractive fragment coverage and extractive fragment density. "c" denotes the compression ratio.
  • Figure 3: Prompt templates and examples for large language models.
  • Figure 4: Automatic evaluation result (ROUGE-2 Score) on CLSum.
  • Figure 5: Automatic evaluation result (BARTScore) on CLSum.
  • ...and 3 more figures