基于法条检索的生成式法律问答研究

李明达; 邸洪波; 孙媛媛; 王艳华; 杨志豪; 林鸿飞

doi:10.13451/j.sxu.ns.2024159

山西大学学报(自然科学版) ›› 2025, Vol. 48 ›› Issue (04) : 653 -665. DOI: 10.13451/j.sxu.ns.2024159

第30届全国信息检索学术会议(CCIR2024)论文选登

基于法条检索的生成式法律问答研究

李明达 ¹ ,
邸洪波 ² ,
孙媛媛 ¹ ,
王艳华 ³ ,
杨志豪 ¹ ,
林鸿飞 ¹

作者信息 +

Research on Generative Legal Question Answering Based on Statutory Articles Retrieval

Author information +

文章历史 +

PDF (3113K)

摘要

基于预训练语言模型的法律问答系统无法灵活理解用户的意图，且缺乏外部知识的整合，难以达到预期的效果。为此，该文提出了基于法条库的细粒度刑法问答数据集（Fine-grained Criminal Law Question Answering,FCL-QA），并基于该数据集，提出了基于大语言模型的法条检索增强问答框架（Statutory Articles Retrieval Augmented Question Answering Framework,SaRAF）。其核心思想是通过多等级分类定位到问题所属的主题，通过主题缩小法条的范围便于进行检索，并最终利用大语言模型生成答案。实验结果表明，SaRAF框架优于无法条生成与传统检索增强生成（Retrieval-augmented Generation,RAG）的方法，在FLC-QA数据集上取得了42.27%的ROUGEL F1分数、27.78%的BLEU-4分数和72.52%的BERTScore分数。

Abstract

Pre-trained language model-based legal question answering systems struggle to flexibly understand users' intent and lack the integration of external knowledge, making it difficult to achieve desired results. To address this, this paper proposes a fine-grained legal question answering dataset based on the criminal law articles library (FCL-QA). Based on FCL-QA, this paper proposes a Statutory Articles Retrieval Augmented Question Answering Framework (SaRAF) based on large language model. The core idea is to locate the category of the question through multi-level classification, narrow the scope of statutory articles through the category to facilitate retrieval, and finally generate the answer using a large language model. Experimental results show that the SaRAF outperforms both without statutory articles generation method and Retrieval-augmented Generation（RAG) method, achieving ROUGE-L F1 score of 42.27%, BLEU-4 score of 27.78% and BERTScore of 72.52% on the FCL-QA dataset.