A Tree-of-Thoughts Inspired Hybrid Approach for Legal Case Judgement Summarization using LLMs

2026-06-26

AI Summary

This work proposes a novel hybrid extractive-abstractive summarization approach for legal judgment texts, inspired by the Tree of Thought (ToT) framework. Addressing the limitations of purely extractive or purely abstractive methods—which often struggle to balance factual accuracy with linguistic fluency—the proposed method introduces, for the first time, a ToT-based reasoning mechanism into legal summarization. It employs a specially designed prompting strategy that synergistically integrates content selection and text generation through structured reasoning paths. Leveraging large language models such as DeepSeek and Llama, the approach enables coherent and accurate summary production by jointly optimizing extraction and generation steps. Experimental results demonstrate that the proposed method significantly outperforms conventional extractive and abstractive baselines in summary quality, thereby validating the effectiveness and potential of hybrid strategies in legal document summarization.

Abstract

In recent times, Large Language Models (LLMs) are increasingly being used for legal case judgement summarization. Most prior works have tried traditional extractive and abstractive summarization of case judgements. However, hybrid or extractive-abstractive techniques have not been explored much. In this work, we propose a novel tree-of-thoughts inspired extractive-abstractive summarization approach for legal judgement summarization. We conduct experiments using two popular LLMs, DeepSeek and LLama, and compare among extractive, abstractive and extractive-abstractive summarization. Our experiments show that the proposed extractive-abstractive prompt provides better summaries compared to other types of LLM prompts.

Problem

• legal case judgement summarization
• extractive-abstractive summarization
• Large Language Models
• hybrid approach

Innovation

• Tree-of-Thoughts
• extractive-abstractive summarization
• legal judgement summarization
• Large Language Models
• hybrid summarization