Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

THUIR@COLIEE 2023: Incorporating Structural Knowledge into Pre-trained Language Models for Legal Case Retrieval

Haitao Li, Weihang Su, Changyue Wang, Yueyue Wu, Qingyao Ai, Yiqun Liu

TL;DR

The paper tackles legal case retrieval in COLIEE 2023 Task 1 by introducing SAILER, a structure-aware pre-trained language model that explicitly models the Fact, Reasoning, and Decision sections of cases. It combines targeted pre-processing, traditional lexical baselines, SAILER representations, and learning-to-rank with post-processing (trial-date filtering, query-case filtering, dynamic cut-off) to fuse heterogeneous signals. Empirical results show that SAILER substantially outperforms baselines and that ensemble post-processing yields the best performance, culminating in a championship run. This work demonstrates that incorporating document structure and carefully crafted post-processing can significantly improve retrieval effectiveness in long, structured legal texts, with practical impact for automated precedent discovery.

Abstract

Legal case retrieval techniques play an essential role in modern intelligent legal systems. As an annually well-known international competition, COLIEE is aiming to achieve the state-of-the-art retrieval model for legal texts. This paper summarizes the approach of the championship team THUIR in COLIEE 2023. To be specific, we design structure-aware pre-trained language models to enhance the understanding of legal cases. Furthermore, we propose heuristic pre-processing and post-processing approaches to reduce the influence of irrelevant messages. In the end, learning-to-rank methods are employed to merge features with different dimensions. Experimental results demonstrate the superiority of our proposal. Official results show that our run has the best performance among all submissions. The implementation of our method can be found at https://github.com/CSHaitao/THUIR-COLIEE2023.

THUIR@COLIEE 2023: Incorporating Structural Knowledge into Pre-trained Language Models for Legal Case Retrieval

TL;DR

The paper tackles legal case retrieval in COLIEE 2023 Task 1 by introducing SAILER, a structure-aware pre-trained language model that explicitly models the Fact, Reasoning, and Decision sections of cases. It combines targeted pre-processing, traditional lexical baselines, SAILER representations, and learning-to-rank with post-processing (trial-date filtering, query-case filtering, dynamic cut-off) to fuse heterogeneous signals. Empirical results show that SAILER substantially outperforms baselines and that ensemble post-processing yields the best performance, culminating in a championship run. This work demonstrates that incorporating document structure and carefully crafted post-processing can significantly improve retrieval effectiveness in long, structured legal texts, with practical impact for automated precedent discovery.

Abstract

Legal case retrieval techniques play an essential role in modern intelligent legal systems. As an annually well-known international competition, COLIEE is aiming to achieve the state-of-the-art retrieval model for legal texts. This paper summarizes the approach of the championship team THUIR in COLIEE 2023. To be specific, we design structure-aware pre-trained language models to enhance the understanding of legal cases. Furthermore, we propose heuristic pre-processing and post-processing approaches to reduce the influence of irrelevant messages. In the end, learning-to-rank methods are employed to merge features with different dimensions. Experimental results demonstrate the superiority of our proposal. Official results show that our run has the best performance among all submissions. The implementation of our method can be found at https://github.com/CSHaitao/THUIR-COLIEE2023.
Paper Structure (27 sections, 13 equations, 2 figures, 5 tables)

This paper contains 27 sections, 13 equations, 2 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Legal Case Retrieval
  4. Dense Retrieval
  5. Task Overview
  6. Task Description
  7. Data Corpus
  8. Metrics
  9. Method
  10. Pre-processing
  11. Remove useless information
  12. Summary extraction
  13. Reference sentence extraction
  14. Traditional Lexical Matching Models
  15. TF-IDF
  16. ...and 12 more sections

Figures (2)

  • Figure 1: An example of the legal case structure in the Case Law system.
  • Figure 2: The model design for SAILER, which consists of a deep encoder and two shallow decoders. The Reasoning and Decision section are aggressively masked, joined with the Fact embedding to reconstruct the key legal elements and the judgment results.