LAW-Chat: Conversational Access to Bharatiya Nyaya Sanhita Using Semantic Retrieval Techniques

View at publisher Summarise with VIRAT
Authors: Aparna Raj, Manas Chandewar, Dhananjay Rupanwar, Navnath Kale
Cited by: 0
References: 5

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Digital solutions for understanding criminal law have become essential with the introduction of the Bharatiya Nyaya Sanhita (BNS). However, complex legal terminology and fragmented statutory content continue to limit accessibility for non-expert users. This paper presents LAW-Chat, a conversational legal assistant that enables users to ask natural-language questions and retrieve relevant BNS provisions with simplified explanations. Unlike keyword-based search systems, LAW-Chat uses transformer-based semantic embeddings stored in a vector database for context-aware retrieval. Experimental evaluation on real-world legal queries achieved an accuracy of 90.3%, significantly outperforming traditional keyword-based approaches (72.5%). The results demonstrate the potential of conversational AI to enhance legal literacy and public access to Indian criminal law.

Fields of Study & Subjects

Artificial Intelligence in Law Topic Modeling Information Retrieval and Search Behavior

Publication details

Journal / source
IEEE International Students' Conference on Electrical, Electronics and Computer Science (SCEECS)
Pages
1–5
Publication date
31 January 2026
Language
EN