Study On The Construction Of Health Knowledge Graph Of Traditional Chinese Medicine

Knowledge graph is an important data resource for knowledge management and application in the era of large data.It has become the key technology of semantic retrieval and knowledge-based reasoning and decision-making in all fields.As an important member of the semantic network,the knowledge graph makes the storage of large-scale knowledge more standardized and more efficient.Knowledge graph often contains various entities and their attributes,as well as the semantic relations between the various entities.The construction of knowledge graph involves many concrete technical aspects,such as named entity recognition,relational extraction,data fusion,knowledge reasoning and knowledge representation,and ontology is the main method of representing the conceptual model of knowledge graph.In the field of Web search and some common areas,a variety of large-scale knowledge graph libraries have been formed,but the construction of knowledge graphs in the field of medicine and traditional Chinese medicine is still in its infancy,although there are some large-scale medical ontology,but there is less research in the construction of knowledge graph on specialized medical specially traditional Chinese medicine,which greatly hindered the application of traditional Chinese medicine knowledge and sharing of knowledge.Therefore,this paper makes symptom,syndromes,disease and medicine as the main entity in the traditional Chinese medicine health knowledge graph construction research through the integration of a variety of data resources,the main contents and results include the following two aspects:(1)To solve the problem of constructing the main traditional Chinese medicine conceptual entities in knowledge graph construction,such as symptoms,syndromes,diseases and traditional Chinese medicine.The corresponding graph Schema is designed to determine the basic category,category attribute and semantic relation of the graph.On this basis,through the processing and integration of four different data sources(including the Baidu Encyclopedia Knowledge Base,Spleen and Stomach Clinical Case Data,Disease Classification Data and the existing Western medicine ontology),we used the information extraction and correlation analysis to extract knowledge from the different data sources,and then using the entity alignment method based on attribute vector to integrate the knowledge of different source data.Then the Traditional Chinese medicine health knowledge graph which contains four kinds of entities(3927 kinds of diseases,2128 kinds of diseases,450 kinds of syndromes and 572 kinds of traditional Chinese medicine)and five kinds of semantic relations were formed.Finally,this paper uses the data generation function of Jena to carry out the knowledge graph in OWL representation and make data generation.(2)Through the Protege ontology editor,this paper adds the constraint definition to the entities and their relations in TCM(Traditional Chinese medicine)knowledge graph,and uses Protege to graphicize some of the knowledge in the knowledge graph.Finally,based on the knowledge graph,the paper uses the open source toolkit Jena and the reasoning rules based on the logic of the TCM diagnosis to carry on the knowledge reasoning demonstration analysis and application.The analysis results show that it has certain feasibility and diagnosis application value.In this paper,the research on TCM knowledge graph construction focuses on the research of knowledge representation and the fusion of multiple data sources,but it needs to be further studied in the application of knowledge reasoning and knowledge learning,which will be improved in the follow-up study.