DEVELOPING AN INTELLIGENT MEDICAL CONSULTATION CHATBOT WITH RAG, USER MEMORY, AND PROACTIVE INTERACTION

Abstract

In recent years, artificial intelligence (AI) has become more popular and widely applied in many fields. However, in Vietnam, the use of AI in healthcare is still quite limited. This thesis focuses on building a smart medical chatbot using large language models (LLMs), aiming to support users with health-related questions. The system combines several technologies, including Retrieval-Augmented Generation (RAG), OpenAI's API, user memory, and proactive interaction features. A vector database is used for document retrieval, and natural language processing (NLP) techniques help improve the chatbot’s responses. Through testing, the chatbot showed good performance in answering user queries in a natural and relevant way. The results suggest that this chatbot model could be useful and has the potential for further development in healthcare applications. Keywords: Artificial Intelligence (AI), Medical Chatbot, Retrieval-Augmented Generation (RAG), Large Language Model (LLM), User Memory, Natural Language Processing (NLP), Vector Database, Proactive Interaction.