JUSTICE: Judicial Unified Synthesis Through Intermediate Conclusion Emulation for Automated Judgment Document Generation

TL;DR

JUSTICE addresses a key gap in automated judgment document generation by explicit modeling of the Pre-Judge phase. It decomposes the process into Referential Judicial Element Retrieval (RJER), Intermediate Conclusion Emulation (ICE), and Judicial Unified Synthesizer (JUS) to mirror the human judge’s workflow: Search → Pre-Judge → Write. Empirical results on JuDGE and LeCaRDv2-Doc show that the framework yields superior legal accuracy and text quality, with notable gains in Convicting and Penalty predictions and higher semantic fidelity. The modular design enables verifiable intermediate conclusions and robust synthesis, highlighting the practical importance of grounding judgments in structured precedents and statutes. The work suggests broader applicability to other legal systems and stresses offline deployment for confidentiality and safety.

Abstract

Automated judgment document generation is a significant yet challenging legal AI task. As the conclusive written instrument issued by a court, a judgment document embodies complex legal reasoning. However, existing methods often oversimplify this complex process, particularly by omitting the ``Pre-Judge'' phase, a crucial step where human judges form a preliminary conclusion. This omission leads to two core challenges: 1) the ineffective acquisition of foundational judicial elements, and 2) the inadequate modeling of the Pre-Judge process, which collectively undermine the final document's legal soundness. To address these challenges, we propose \textit{\textbf{J}udicial \textbf{U}nified \textbf{S}ynthesis \textbf{T}hrough \textbf{I}ntermediate \textbf{C}onclusion \textbf{E}mulation} (JUSTICE), a novel framework that emulates the ``Search $\rightarrow$ Pre-Judge $\rightarrow$ Write'' cognitive workflow of human judges. Specifically, it introduces the Pre-Judge stage through three dedicated components: Referential Judicial Element Retriever (RJER), Intermediate Conclusion Emulator (ICE), and Judicial Unified Synthesizer (JUS). RJER first retrieves legal articles and a precedent case to establish a referential foundation. ICE then operationalizes the Pre-Judge phase by generating a verifiable intermediate conclusion. Finally, JUS synthesizes these inputs to craft the final judgment. Experiments on both an in-domain legal benchmark and an out-of-distribution dataset show that JUSTICE significantly outperforms strong baselines, with substantial gains in legal accuracy, including a 4.6\% improvement in prison term prediction. Our findings underscore the importance of explicitly modeling the Pre-Judge process to enhance the legal coherence and accuracy of generated judgment documents.

Figures (10)

• Figure 1: Illustration of the challenges of the existing Judgment Document Generation method.
• Figure 2: Main framework of JUSTICE.
• Figure 3: Impact of the ICE module on judgmental accuracy.
• Figure 4: Impact of retrieved article quantity ($k_2$).
• Figure 5: Impact of different reference judicial element types.