基于中医网络的中心性算法研究

Centrality Research Based on Traditional Medicine Network

中医概念之间的关系极为复杂,采用传统的基于简单数据集的数据挖掘方法显得力不从心。考虑到中医数据的特点,提出了一种基于图结构的挖掘方法,将中医对象之间的复杂关系和潜在的信息提取出来。首先,通过自然语言处理从医案中抽取出的中医概念与中医本体知识库匹配得到中医的知识网络。然后将该中医网络抽象成数学表达方式——图,利用图论的算法来处理。最后,采用中心性算法来分析中医网络,找出在中医诊断网络中具有重要作用的症状。

The relationships between TCM （traditional Chinese medicine） concepts are so complex that conventional data mining based on simple data set cant be capable of processing. Considering the property of TCM, a mining method based on graph data is proposed, which is used to extract complex relationships and potential information among TCM objects. To begin with, the TCM knowledge network is acquired by matching TCM concepts which are extracted from medicine cases through NLP and TCM ontology knowledge base. Then, the TCM network is transformed into mathematical expression - -graph, which is processed according to graph algorithms. Finally, centrality arithmetic is adopted to analyze traditional Chinese medicine network to find out the symptoms which play a significant role in the network analysis.