An LLM Powered Legal Data Warehouse and Precedent Retrieval System for Reducing Judicial Backlog
International Journal of Science and Research (IJSR) · 30.06.2026 · 1613-1615
DOI: 10.21275/sr26629132934 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Science and Research (IJSR), erschienen am 30.06.2026. Verfasst von Debopriyo Sarkar und 1 weiteren Autor:innen.
Abstract (Originalsprache)
: Judicial systems globally face escalating backlogs, primarily driven by the time intensive nature of legal research and precedent analysis. This paper introduces the Juris Precedent Retrieval & Synthesis Model (JPRS Model), an LLM powered architecture designed to bridge the gap between historical judicial data and active case resolution. By extracting and synthesizing massive volumes of historical rulings into a structured legal data warehouse, the JPRS Model provides judges with a high fidelity 'decision support' framework. Our evaluation demonstrates that the model's core components the Precedent Summarization Module (PRE-SUM) and the Judicial Recommendation Engine (JUR-REC) significantly streamline preliminary case scrutiny, yielding a 38% reduction in review time. Ultimately, this framework ensures that judicial professionals can navigate decades of complex legal reasoning efficiently without sacrificing the refined context of historical rulings .
Zitieren
APA
Sarkar, D., & Tewary, S. (2026). An LLM Powered Legal Data Warehouse and Precedent Retrieval System for Reducing Judicial Backlog. International Journal of Science and Research (IJSR), 1613-1615. https://doi.org/10.21275/sr26629132934
Vancouver
Sarkar D, Tewary S. An LLM Powered Legal Data Warehouse and Precedent Retrieval System for Reducing Judicial Backlog. International Journal of Science and Research (IJSR). 2026:1613-1615. doi:10.21275/sr26629132934
BibTeX
@article{Sarkar2026,
author = {Sarkar, Debopriyo and Tewary, Sunetra},
title = {An LLM Powered Legal Data Warehouse and Precedent Retrieval System for Reducing Judicial Backlog},
journal = {International Journal of Science and Research (IJSR)},
year = {2026},
pages = {1613--1615},
doi = {10.21275/sr26629132934}
}
EUVIMED