A Medical Answering System Based On Knowledge Graph

With the development of the Internet,the demand of medical knowledge search has increased sharply,but the current medical knowledge websites are too professional to make users find what they want to know.Also because of the lack of pertinence,the websites' search engine can't take different people's situations into consideration.In order to alleviate this contradiction,by using knowledge graph technology,this paper designed and developed a medical expert support system based on knowledge graph.With Natural Language Processing technology,extracting knowledge from the unstructured content of electronic medical records,and fuse them into knowledge graph.Based on the knowledge graph,on the technology of semantic search and question answering system,the system also provides medical semantic search and question answering service.The system can directly understand users' intentions,so that users do not need to find information they need in professional websites,and at the same time,it can make targeted answers according to different users' input.The contribution of this paper includes:1)Design and implement a medical knowledge graph building system for unstructured data,which is detailed from three aspects,named entity recognition,relation extraction and knowledge fusion.2)Design and implement a medical semantic search and answering model which is detailed in three aspects:data sets construction,algorithm evaluation and results presentation.3)Design and implement a medical answering system based on knowledge graph.It provided medical naming entity identification,medical relation extraction,knowledge graph visualization,medical semantic search and medical question answering service.