A RAG-based approach for medical question answering using knowledge graphs and pretrained models

With the rapid development of artificial intelligence technologies, medical question-answering systems that integrate knowledge graphs and large models have shown great potential in the field of smart healthcare. This paper proposes a medical question-answering system based on a Knowledge Graph and Retrieval-Augmented Generation (RAG) framework. The system combines Named Entity Recognition (NER) and intent recognition techniques to perform reasoning using the knowledge graph to generate accurate medical answers. The system first preprocesses medical texts and applies rule-based classification to train the intent classification and NER models, which then extract relevant information from the knowledge graph to generate query statements. Based on the generated queries, the system combines the user's query with the knowledge graph retrieval results and uses a large model to generate the final answer. Experimental results show that the system can provide efficient and accurate answers for medical question-answering tasks, especially in specific medical domains, demonstrating good applicability.