Analysis And Design Of Health Question Answering System Based On Knowledge Graph

With the improvement of people’s living standard and the implementation of national smart medical policy,medical and health has become the focus of public attention,and medical and health data on the internet is becoming more and more abundant.However,it is difficult for people to get the information they need in a timely manner from the huge amount of medical and health data.Therefore,it is of great theoretical and practical significance to build an interactive question answering system based on knowledge graph to assist medical treatment.In this dissertation,we first construct a knowledge graph for the medical and health field,then design a health question answering system based on knowledge graph,and finally implement and test the health question answering system.The main research contents of this dissertation are as follows:(1)Construction of knowledge graph in medical and health field.Firstly,the crawler technology is used to obtain medical data from medical professional websites.Secondly,the data is cleaned by setting stop words and maximum forward/backward matching algorithm.Then the entities,relationships and attributes are extracted.Finally,the knowledge graph of medical and health field is stored in Neo4 j with44112 entities and 291164 relations.(2)Design of health question answering system based on knowledge graph.It mainly includes entity extraction,intention classification and system answer of user medical question.Firstly,Aho-Corasick automation algorithm and semantic similarity calculation method are used to extract the entity in the user question.Secondly,the intention of user question is classified by using the classification method based on feature words and Naive Bayesian.Finally,the entity and intention are parsed into Cypher statement,which is used to conduct knowledge query in the knowledge graph to get results.(3)Implementation of health question answering system.A medical and health information retrieval platform is provided for users.The platform of the health question answering system is mainly implemented by the Flask framework,and the front-end knowledge graph is visualized by the visualization tool D3.Finally,the results are feedback to users in the form of text and visual knowledge graph.In summary,this dissertation implements a health question answering system based on knowledge graph,which can respond to users’ questions in a timely manner,present results in the form of text and visual knowledge graph.It assists users in preliminary self-diagnosis,and to a certain extent,also reduces the pressure of doctors.