Research On Health Medical Knowledge Push System Based On Knowledge Graph

People often do not go to the hospital without serious illness pain,but hope that through the search engine or online medical inquiry system to obtain information of their symptoms,while the existing medical QA system is too dependent on the doctor's personal experience,does not conform to the scientific concept of medical data driven.The development of information technology,such as large data,artificial intelligence and Internet of things,makes the integration of medical industry and computer information technology inevitable,the knowledge map provides a method to analyze the relationship between "things" point of view,Knowledge map based health knowledge push system can predict users' diseases according to the symptoms of users,so as to push medical knowledge related to patients' diseases.First,the structural features of the electronic medical records are analyzed,the main source of knowledge to construct the knowledge map is the first course record and discharge summary in the electronic medical record,then we put forward the medical entity extraction task of the electronic medical record,and use the CRF model as the method of extracting the medical entity,and annotate some electronic medical records,Then use the CRF++ tool to extract the medical entity in the Chinese electronic medical record,and convert the extracted entity into the form of the SOP triple.Secondly,aiming at how to calculate the influence factor IMPT of a sick entity for a disease entity,naive Bayes algorithm is used to calculate the IMPT value of sick entities.After that,the constraints of entity nodes are set up.Then,the weight factor of medical entity and disease is imported into Neo4j by Neo4j Jdbc,and the knowledge map is constructed.Finally,the system structure design and technical route analysis are carried out on the medical knowledge push system,using Spring Boot to integrate Spring Data Neo4j and Spring Data JPA to complete the system framework construction,a disease judgment strategy based on knowledge atlas is proposed in the realization of the system function,And according to the formulation of the disease judgment strategy,choose the most likely disease and other possible disease of the user to recommend to the user,Finally,according to all the diseases that the patient may have,provide the corresponding medical knowledge for the patient.This paper uses knowledge Graph as the core technology,through the query on the user input symptoms in the knowledge map,and according to the query results to provide users with disease related medical knowledge,the experimental results show that the medical knowledge push system for users is helpful to diagnosis of certain diseases...