MultiLCR studies legal case retrieval beyond case-level textual similarity by modeling defendants' roles and judgment outcomes. It introduces the Multi-ELAM and Multi-LeCaRDv2 datasets with defendant-level annotations and a role- and judgment-aware retrieval framework.

Publication

Accepted to the EMNLP 2026 main conference and published in the conference proceedings (October 2026).

ACL Anthology ID: 2026.emnlp-main.1918. Pages: 43688–43713. Publisher: Association for Computational Linguistics. DOI: 10.18653/2026.emnlp-main.1918.

Overview

Surface-level textual matching can overlook the distinct roles and judgment outcomes of defendants in multi-defendant cases. The paper enriches existing benchmarks with defendant-level annotations and proposes a role- and judgment-aware framework. Its abstract reports improvements over state-of-the-art baselines for multi-defendant legal case retrieval.

Resource Availability

  • The Multi-LCR dataset is available on Hugging Face.
  • See the GitHub repository for code availability and setup instructions.

Resources