RLJP: Legal Judgment Prediction via First-Order Logic Rule-enhanced with Large Language Models

2025-05-27

AI Summary

Existing legal judgment prediction models either neglect legal reasoning logic or rely on rigid logical rules, rendering them ill-suited to the dynamic reasoning structures of complex cases. To address this, we propose an adaptive judgment reasoning framework integrating first-order logic (FOL) and contrastive learning. Its core innovation is Confusion-Aware Contrastive Learning (CACL), which implements a human-like, three-stage process: rule initialization → confusion-case-driven dynamic optimization → judgment prediction. The method jointly leverages large language models for semantic understanding, judicial precedents, and domain-specific legal knowledge, with reasoning formalized in FOL for interpretability and adaptability. Evaluated on two public legal judgment datasets, our approach consistently outperforms state-of-the-art methods across all metrics. Source code is publicly available.

Technology Category

Machine Learning: Statistical Relational/Logic LearningCognitive Modeling & Cognitive Systems: Conceptual Inference and ReasoningKnowledge Representation and Reasoning: Computational Complexity of Reasoning

Application Category

Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for rankingSemantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMsUser Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Legal Judgment Prediction (LJP) is a pivotal task in legal AI. Existing semantic-enhanced LJP models integrate judicial precedents and legal knowledge for high performance. But they neglect legal reasoning logic, a critical component of legal judgments requiring rigorous logical analysis. Although some approaches utilize legal reasoning logic for high-quality predictions, their logic rigidity hinders adaptation to case-specific logical frameworks, particularly in complex cases that are lengthy and detailed. This paper proposes a rule-enhanced legal judgment prediction framework based on first-order logic (FOL) formalism and comparative learning (CL) to develop an adaptive adjustment mechanism for legal judgment logic and further enhance performance in LJP. Inspired by the process of human exam preparation, our method follows a three-stage approach: first, we initialize judgment rules using the FOL formalism to capture complex reasoning logic accurately; next, we propose a Confusion-aware Contrastive Learning (CACL) to dynamically optimize the judgment rules through a quiz consisting of confusable cases; finally, we utilize the optimized judgment rules to predict legal judgments. Experimental results on two public datasets show superior performance across all metrics. The code is publicly available{https://anonymous.4open.science/r/RLJP-FDF1}.

Problem

• Enhancing Legal Judgment Prediction with adaptive logic rules
• Addressing rigidity in existing legal reasoning models
• Improving performance in complex, lengthy legal cases

Innovation

• Uses first-order logic for legal reasoning
• Applies confusion-aware contrastive learning
• Dynamically optimizes judgment rules