Medical Knowledge Graph Construction And Intelligent Question Answering Research

With the advancement of the times and the development of science and technology,technology is gradually changing our lives.Today,artificial intelligence technologies such as machine learning and deep learning are constantly being updated and developing rapidly.Knowledge graphs that represent structural relationships between entities and intelligent question answering systems that can acquire information accurately and quickly and give answers have become more and more popular research directions in artificial intelligence.Real life contains a variety of relationships,how to make full use of the knowledge graph to mine and represent the relationship between information becomes particularly important.A knowledge graph is essentially a semantic network that can reveal relationships between entities.The main purpose of the research in this paper is to make the constructed knowledge graph more effective,fast,accurate and to form a certain scale.An intelligent question answering system is essentially an advanced form of information retrieval system.The emergence of deep learning technology has opened up new ways for the establishment and research of knowledge graphs and intelligent question answering systems,and greatly promoted the research and development of knowledge graphs and intelligent question answering systems.This paper mainly conducts in-depth mining based on professional knowledge in the medical field,constructs a large-scale Chinese medical knowledge graph based on the information relationship between medical knowledge,and implements intelligent question answering on the knowledge graph.Conduct research on knowledge graphs and intelligent question answering systems.The main research contents of this paper include the following three aspects:1.Construction of medical knowledge graph: In view of the fact that the number of large-scale medical knowledge graphs in China is relatively small,this paper conducts research on it.First,crawler technology is used to crawl popular medical knowledge websites in China,mainly for unstructured data,semi-structured data The structured data and structured data are collected and preprocessed,and the preprocessed data is subjected to entity recognition and relationship extraction.This paper adopts the joint extraction model based on Bi Lstm+Attention.The comparative experiments show that the F1 score of the joint extractor in this paper is 86.2%,which is much higher than other models.After passing through the joint extractor,the medical entities,inter-entity relationships and the entity’s own attributes are obtained.It performs word similarity calculation and knowledge fusion of multiple knowledge bases to obtain triples and form knowledge graphs and store them in Neo4 j and Mongo DB databases.Finally,a Chinese medical knowledge map is formed,including 8807 diseases,4982 medicines,3519 diagnosis and inspection items,6919 foods,54 medical departments,15023 sales medicines and 4901 disease symptoms,forming a total scale including Knowledge graph of 44205 medical entity information.2.Intelligent question answering model: Since the methods of traditional intelligent question answering systems are mostly based on the form of rule matching,in view of the shortcomings of traditional intelligent question answering systems such as slow speed and poor effect,this paper proposes a model based on dynamic memory network combined with attention mechanism.Accurately grasp the questions raised by users,and the model can update the internal information in real time to improve the accuracy of the model.Experiments show that the accuracy of the model can reach 88.7%,which is the best in comparison experiments with various models.3.Web development of medical knowledge graph and intelligent question answering system: Based on the constructed medical knowledge graph,this article applies it to the intelligent question answering model.In order to provide users with a good experience,this article develops the system on the web,and mainly realizes two major aspects.One is to realize the intelligent question and answer function that interacts with users,so that users can get accurate answers by chatting.The second is the visualization of knowledge graphs,which presents users with structured and highly correlated knowledge graphs.Users only need to enter a few keywords to intuitively see the knowledge graphs displayed in the form of pictures.The front-end and back-end separation technology used in the development is developed using the Flask framework of the python language.After testing,it is found that the system has good stability and accuracy,and can meet the user’s use standard.