基于框架语义的在线医疗评论情感分析  

Sentiment Analysis of Medical Online Comments Based on Frame Semantics

作　　者：由丽萍[1] 何玲玲 You Liping;He Lingling(School of Economics and Management,Shanxi University,Taiyuan 030006,China)

机构地区：[1]山西大学经济与管理学院

出　　处：《现代情报》2020年第3期111-116,125,共7页Journal of Modern Information

基　　金：国家自然科学基金项目“基于FrameNet的中文评价词汇本体构建与观点挖掘研究”(项目编号:71403154)

摘　　要：[目的/意义]随着"互联网+"在医疗服务行业的应用与发展,积累了大量的医疗评价信息,利用情感分析技术可以对其进行有效地挖掘和利用,从而为医疗管理提供决策参考。[方法/过程]基于框架语义理论建立医疗情感语义分类词典;采用词典和规则相结合的方法进行在线医疗评论的情感语义分析,标注情感类别、情感主题、极性和强度等信息。[结果/结论]通过在线医疗评论数据测试,验证了研究方法的有效性和科学性,是情感分析向医疗健康领域纵深发展的一次有益探索。[Purpose/Significance]Since a lot of medical evaluation information is accumulated with the application and development of"Internet+"in the medical service industry,sentiment analysis technology can be used to mine and utilize it effectively,so as to provide reference for medical management.[Method/Process]A medical sentiment semantic classification dictionary was constructed based on the Frame Semantics theory.Based on the dictionary and some rules,the sentiment of online medical reviews was analyzed with annotation of sentiment type,sentiment topic,polarity and strength,etc.[Result/Conclusion]Through the test of online medical review data,a higher accuracy rate and recall rate were obtained,which proved the validity and scientificity of the research method.It is a useful exploration for the further development of emotional analysis in the field of medical health.

关 键 词：框架语义 情感分析 医疗信息 在线医疗评论 

分 类 号：F713.5[经济管理—市场营销] G252[经济管理—产业经济]