Intelligent Diagnosis and Treatment System Based on Medical Knowledge Graph

In this paper, RotatE model is trained by medical knowledge base to realize triplet extraction and inference of complex medical knowledge respectively. Based on these two models, this paper constructs a disease-centered medical knowledge map. At the same time, an intelligent medical system is developed based on the constructed knowledge graph. The system can diagnose diseases and deduce complications. The intelligent medical system extracts the triplet of symptoms input by patients through the TPLinker model, and standardizes the corresponding text. In the process of disease diagnosis, based on the RotatE model, the system deduces the relationship between multiple “disease” entities in the medical knowledge graph, which can not only predict the disease suffered by patients, but also deduce the possible complications caused by the disease. This provides theoretical methods and practical examples for the realization of intelligent medicine.