Distinguish Confusion in Legal Judgment Prediction via Revised Relation Knowledge

TL;DR

This paper introduces D-LADAN, an end-to-end model that improves legal judgment prediction by distinguishing confusing law articles using a graph-based approach and a dynamic memory mechanism, addressing semantic similarity and data imbalance issues.

Contribution

The paper proposes D-LADAN, a novel model combining graph distillation and a momentum-updated memory to better differentiate law articles in legal judgment prediction.

Findings

D-LADAN outperforms existing methods in accuracy.

The model demonstrates enhanced robustness.

It effectively handles semantic similarity and data imbalance.

Abstract

Legal Judgment Prediction (LJP) aims to automatically predict a law case's judgment results based on the text description of its facts. In practice, the confusing law articles (or charges) problem frequently occurs, reflecting that the law cases applicable to similar articles (or charges) tend to be misjudged. Although some recent works based on prior knowledge solve this issue well, they ignore that confusion also occurs between law articles with a high posterior semantic similarity due to the data imbalance problem instead of only between the prior highly similar ones, which is this work's further finding. This paper proposes an end-to-end model named \textit{D-LADAN} to solve the above challenges. On the one hand, D-LADAN constructs a graph among law articles based on their text definition and proposes a graph distillation operation (GDO) to distinguish the ones with a high prior semantic similarity. On the other hand, D-LADAN presents a novel momentum-updated memory mechanism to dynamically sense the posterior similarity between law articles (or charges) and a weighted GDO to adaptively capture the distinctions for revising the inductive bias caused by the data imbalance problem. We perform extensive experiments to demonstrate that D-LADAN significantly outperforms state-of-the-art methods in accuracy and robustness.