The Core Drugs Analysis based on Social Network Analysis about Traditional Chinese Medicine Records Semantic Relation

Due to complex and abstract of the relationship for traditional Chinese medicine, conventional analytical methods are difficult to draw potential rules, and social network analysis is the method to solve the problem with complex relationship and structure, which provides an opportunity to use social network analysis approach to analyze medical record prescriptions. In this paper, various traditional Chinese medicine concepts, which include disease, disorder, parties, drugs, therapies, syndromes of the prescriptions, are extracted from the medicine records and turned into the relation graph by using the knowledge of graph theory. At the same time three centrality analysis algorithms including degree centrality, betweenness centrality and closeness centrality are implemented and improved based on the characteristics of traditional Chinese medical records. The core prescription is extracted based on the analysis of the word level and semantic level (drugs Tropism of taste) by using the algorithms above.