A Comprehensive Survey on Legal Summarization: Challenges and Future Directions

TL;DR

This survey addresses the problem of automatic legal document summarization by systematically reviewing over 120 transformer-era studies across datasets, models, and evaluation methods. It maps region-specific challenges, summarization strategies (extractive, abstractive, and hybrid), and methodological approaches, highlighting trends and gaps. Key contributions include a taxonomy of datasets and region-specific datasets, a synthesis of evaluation practices, and a forward-looking agenda emphasizing personalization, multilingual and multimodal summarization, and better metrics. The work provides a foundation for researchers and practitioners to advance reliable, domain-aware legal summarization with practical impact on legal research and decision-making.

Abstract

This article provides a systematic up-to-date survey of automatic summarization techniques, datasets, models, and evaluation methods in the legal domain. Through specific source selection criteria, we thoroughly review over 120 papers spanning the modern `transformer' era of natural language processing (NLP), thus filling a gap in existing systematic surveys on the matter. We present existing research along several axes and discuss trends, challenges, and opportunities for future research.

Figures (5)

• Figure 1: Schematic Legal Summarization Pipeline: Legal summarization pipelines process lengthy, structurally diverse documents to generate summaries, serving both general audiences and domain experts.
• Figure 2: Countries identified in the collection of papers during the survey study
• Figure 3: Legal summarization research trends for last 5 years
• Figure 4: Taxonomy of legal summarization strategies categorized into three main approaches, each with more specific sub-categories.
• Figure 5: Overview of key aspects and dimensions in human evaluation for legal summarization