Incorporating Precedents for Legal Judgement Prediction on European Court of Human Rights Cases

TL;DR

This paper explores methods to incorporate legal precedents into judgment prediction models for European Court of Human Rights cases, improving accuracy especially for less common articles by joint training and strategic integration.

Contribution

It introduces a novel approach combining a fine-grained precedent retriever with joint training of the retriever and judgment prediction models, enhancing performance over previous methods.

Findings

Precedent integration during training improves prediction accuracy.

Joint training of retriever and predictor reduces latent space divergence.

Method benefits sparse legal articles more significantly.

Abstract

Inspired by the legal doctrine of stare decisis, which leverages precedents (prior cases) for informed decision-making, we explore methods to integrate them into LJP models. To facilitate precedent retrieval, we train a retriever with a fine-grained relevance signal based on the overlap ratio of alleged articles between cases. We investigate two strategies to integrate precedents: direct incorporation at inference via label interpolation based on case proximity and during training via a precedent fusion module using a stacked-cross attention model. We employ joint training of the retriever and LJP models to address latent space divergence between them. Our experiments on LJP tasks from the ECHR jurisdiction reveal that integrating precedents during training coupled with joint training of the retriever and LJP model, outperforms models without precedents or with precedents incorporated only at inference, particularly benefiting sparser articles.