AgentsCourt: Building Judicial Decision-Making Agents with Court Debate Simulation and Legal Knowledge Augmentation

TL;DR

This paper introduces AgentsCourt, a multi-agent framework for judicial decision-making that simulates court processes, supported by a large legal knowledge base and benchmark, achieving significant improvements in legal article generation.

Contribution

The paper presents a novel multi-agent framework, SimuCourt benchmark, and Legal-KB knowledge base to enhance judicial decision-making with large language models.

Findings

Outperforms existing methods in legal article generation

Achieves 8.6% and 9.1% F1 score improvements in two settings

Demonstrates effective simulation of judicial decision processes

Abstract

With the development of deep learning, natural language processing technology has effectively improved the efficiency of various aspects of the traditional judicial industry. However, most current efforts focus on tasks within individual judicial stages, making it difficult to handle complex tasks that span multiple stages. As the autonomous agents powered by large language models are becoming increasingly smart and able to make complex decisions in real-world settings, offering new insights for judicial intelligence. In this paper, (1) we propose a novel multi-agent framework, AgentsCourt, for judicial decision-making. Our framework follows the classic court trial process, consisting of court debate simulation, legal resources retrieval and decision-making refinement to simulate the decision-making of judge. (2) we introduce SimuCourt, a judicial benchmark that encompasses 420 Chinese judgment documents, spanning the three most common types of judicial cases. Furthermore, to support this task, we construct a large-scale legal knowledge base, Legal-KB, with multi-resource legal knowledge. (3) Extensive experiments show that our framework outperforms the existing advanced methods in various aspects, especially in generating legal articles, where our model achieves significant improvements of 8.6% and 9.1% F1 score in the first and second instance settings, respectively.