부산대학교 도서관

There are a vast number of legal documents pertaining to various sectors in India including India’s Food Safety Regulations which is governed by the Food Safety and Standards Authority of India (FSSAI). For the individuals without legal expertise, understanding the context of these documents would be challenging. To address this issue, a Retrieval Augmented Generation (RAG) chatbot specially tailored to assist the common people in understanding the information provided in Food Safety Documents is introduced. The proposed work is developed with the emerging advancements in Large Language Models (LLMs) like Generative Pre-trained Transformers (GPTs) and Llama3. The system provides a user - friendly interface enabling users to get clarifications either by raising query or by requesting a summary of a specific section in the document. Crucially, the chatbot operates in both Tamil and English, ensuring accessibility and ease of understanding for users preferring their native language. This bridges the gap between legal complexities and public comprehension while leveraging the importance of linguistic accessibility.