Abstract: Aiming at the challenges of multi-source heterogeneous data and information fragmentation across digital platforms in tunnel disease treatment, an intelligent Q&A system based on a knowledge graph is presented for the tunnel disease management knowledge base. The system utilizes knowledge graph technology and Neo4j graph database technology to achieve user question classification and intention recognition using the naive Bayesian model. The system enables efficient retrieval and precise answers to queries related to tunnel disease management. Experimental results demonstrate its strong performance, practicality, and suitability for interactive question answering in the field of tunnel disease management.

Key words: knowledge graph, tunnel disease treatment, tunnel inspection, Q&, A system