
This paper introduces a novel method for automatically constructing large-scale, high-quality datasets of synthetic query-candidate pairs to enhance the performance of legal case retrieval (LCR) systems, particularly in asymmetric retrieval scenarios where user queries are short and concise.


coremsg

enhancing-legal-case-retrieval-by-automatically-generating-large-scale-datasets-of-synthetic-query-candidate-pairs


Enhancing Legal Case Retrieval by Automatically Generating Large-Scale Datasets of Synthetic Query-Candidate Pairs


title_rewrite


The authors propose an approach to enhance the understanding of case relevance in legal case retrieval, integrating lexical matching, semantic retrieval, and learning-to-rank techniques, along with heuristic pre-processing and post-processing methods.


enhancing-legal-case-retrieval-through-improved-understanding-of-case-relevance


Enhancing Legal Case Retrieval through Improved Understanding of Case Relevance



Introducing DELTA, a discriminative model for legal case retrieval, focusing on key facts for improved relevance determination.


delta-pre-train-a-discriminative-encoder-for-legal-case-retrieval-via-structural-word-alignment


DELTA: Pre-train a Discriminative Encoder for Legal Case Retrieval via Structural Word Alignment
