基于“种子概念”的中医医案症状信息提取方法  

A Method for Information Extraction of TCM Symptoms Based on Seed Concept

作　　者：徐亮[1] 陈阳 陈守强[1] 左瑶瑶 毕思玲 袁锋[3] Xu Liang;Chen Yang;Chen Shouqiang;Zuo Yaoyao;Bi Siling;Yuan Feng(The Second Hospital Affiliated to Shandong University of Traditional Chinese Medicine,Jinan 250001,China;School of Information Science and Engineering of Shandong Normal University,Jinan 250001,China;Key Laboratory of TCM Data Cloud Service in Universities of Shandong/Shandong Management University,Jinan 250001,China)

机构地区：[1]山东中医药大学第二附属医院,山东济南250014 [2]山东师范大学信息科学与工程学院,山东济南250014 [3]山东省高等学校中医药数据云服务重点实验室/山东管理学院,山东济南250001

出　　处：《亚太传统医药》2018年第9期112-114,共3页Asia-Pacific Traditional Medicine

基　　金：国家社会科学基金(No:16BGL181)

摘　　要：目的:基于种子概念提取方法,建立中医医案症状信息提取方法。方法:对中医医案进行分词,通过设立种子概念,统计互信息量,选择大于阈值者为复合种子概念;结合萤火虫算法,统计相邻词语间关联性,获取扩展种子概念。结果:获取复合种子概念217个,扩展种子概念68个。结论:该方法实现了中医医案症状信息的自动化提取,为中医医案的数据挖掘提供了便利。Objective：To establish the information extraction method for TCM symptoms based on seed concept.Methods：Firstly,segment words of TCM records,and set the concepts of ＂seed＂;secondly,calculate the mutual information,and select the concepts above threshold as composite seed concept;finally,get the extended concepts by calculating the correlation between adjacent words based on firefly algorithm.Results：We got 217 composite seed concepts and 68 extended concepts.Conclusion：The method realize the automatic extraction of TCM symptoms information,which is helpful for data mining of TCM records.

关 键 词：中医医案 症状信息 信息提取 种子概念 萤火虫算法 

分 类 号：R249[医药卫生—中医临床基础]