Building an Information Retrieval System for University Documents Based on Generative AI Technologies

Abstract

This study presents the development of a Generative AI-Based Information Retrieval System tailored for university documentation. The system leverages cutting-edge technologies, including LangChain, Facebook AI Similarity Search (FAISS) vector search, and OpenAI’s API, to facilitate semantic search and natural language processing (NLP), ensuring efficient and accurate document retrieval. The backend is built with Django, while the frontend leverages Angular for a responsive and user-friendly interface. The system's architecture ensures scalability and high retrieval accuracy through a Retrieval-Augmented Generation (RAG) model. Challenges faced include optimizing system performance, managing dependency conflicts across different libraries, and ensuring secure authentication mechanisms. This paper discusses the design, implementation, and evaluation of the system, highlighting its impact on academic document accessibility and future research directions.