Research On The Construction And Application Of Traditional Chinese Medicine Formula Knowledge Graph

Traditional Chinese medicine is the treasure of traditional Chinese culture and the crystallization of clinical diagnosis and treatment thoughts of ancient doctors.In recent years,the scientific research and clinical value of traditional Chinese medicine has been gradually discovered by the world.Traditional Chinese medicine prescriptions carry thousands of years of medical knowledge and clinical experience of ancient ancestors.A large number of ancient Chinese medicine documents,such as Synopsis of the Golden Chambers and Emergency Prescriptions for the Backup of the Elbows,record the relevant contents of traditional Chinese medicine clinical prescriptions.After thousands of years of medical testing and improvement,they have been proved to have great medical research value and important clinical significance.Under the background of big data and artificial intelligence,the application and rediscovery of the knowledge of traditional Chinese medicine is one of the important research works for the inheritance and innovative development of traditional Chinese medicine.With the advent of smart + medical wave,more and more people are involved in the research of traditional Chinese medicine.Unfortunately,there is no unified standard for the names of many medicinal materials,diseases and syndromes.As a result,many different descriptions are derived,so that TCM workers may need to spend a lot of time on comparison and judgment.In order to solve the confusion of terms that users may encounter in the process of using the data of traditional Chinese medicine prescriptions,so that users can obtain information more accurately,this paper attempts to construct knowledge graph by combining the thesaurus,standardize various TCM entities,and then organize and mine the information of traditional Chinese medicine prescriptions.And supplement and improve the professional control word list and term system to help realize the digitalization of traditional Chinese medicine knowledge.The content of this article contains three parts:First of all,this paper defines the relevant definitions,including knowledge graph and thesaurus,and generally expounds the development of TCM knowledge graph and thesaurus of TCM terms at home and abroad,which lays a research foundation for further development of the following chapters.Secondly,the knowledge graph of traditional Chinese medicine prescriptions is constructed.In the section of entity recognition of traditional Chinese medicine prescriptions,the BERT-Bi LSTM-CRF model used in this paper,its working principle and the selection criteria of traditional Chinese medicine prescriptions data set are introduced in detail,and the text annotation work of the field of functional indications is explained in detail.Finally,the evaluation criteria of the model is explained through confusion matrix.In the section of Entity Relationship Definition,several entity relationship extraction methods are introduced in detail,and how to define entity relationship in this paper.In the knowledge storage section,entities and relationships in traditional Chinese medicine prescription data are stored and visualized through Neo4 j graph database,and the knowledge graph of traditional Chinese medicine prescription is constructed,which supports the subsequent perfect application of knowledge graph in the form of knowledge base.Finally,human-computer interactive retrieval is introduced.In order to further improve and utilize the graph,this paper combined the interactive retrieval on the basis of the knowledge graph in order to improve the retrieval efficiency.Interactive retrieval,query extension,relevant feedback and related concepts and technologies are explained in detail.In view of the problems such as chaotic use of TCM terms and slow updating of standards,which tasks should be realized in human-computer interactive retrieval,a human-computer interactive retrieval framework based on the knowledge graph of TCM ancient prescriptions is designed on this basis,and the concept and function of each part of the framework is explained in detail.Combined with the case of the novel coronavirus,the retrieval effect after the introduction of human-computer interactive retrieval is specifically demonstrated,and the superiority of human-computer interaction is proved by comparing the query in general circumstances.At the end of the chapter,the application realization of knowledge graph of traditional Chinese medicine prescriptions combined with human-computer interactive retrieval is explained.