基于观点挖掘的突发事件微博意见领袖识别方法  

A Microblog Opinion Leader Identification Method Based on Opinion Mining in Emergencies

作　　者：刘高勇[1] 谭依雯 艾丹祥[1] 黄靖钊 Liu Gao-yong;Tan Yi-wen;Ai Dan-xiang;Huang Jing-zhao(School of Management,Guangdong University of Technology,GuangZhou 510520,China)

机构地区：[1]广东工业大学管理学院,广东广州510520

出　　处：《广东工业大学学报》2021年第4期41-51,共11页Journal of Guangdong University of Technology

基　　金：国家自然科学基金资助项目(71740024);广东省哲学社会科学“十三五”规划2020年度课题(GD20CTS01);广州市哲学社会科学发展“十三五”规划2018年度课题(2018GZYB67)。

摘　　要：基于意见领袖概念的本质,运用观点挖掘技术研究突发事件中微博意见领袖的识别,为网络舆情治理提供参考。提出三步识别方法框架:首先采用文献分析法构建指标模型,评价微博博主的信息影响力;其次构建文本主客观分类模型,计算高影响力博主事件相关博文的观点输出性,识别观点博文;然后针对观点博文的评论文本构建情感极性分类模型,计算博文观点获得的支持度以及博主观点的支持率,最终将输出了观点并获得较多支持的高影响力博主作为该事件的“意见领袖”。同时,运用上述方法对典型实际案例的微博舆情数据进行分析,识别该事件的微博意见领袖,对其特征和舆情参与行为进行观察。并将该结果与社会网络分析法、专家人工分析法识别的结果进行了对比,验证了本方法的科学性和有效性。结果表明,本方法能够识别在突发事件舆情生命周期中真正拥有舆论引导力,具备治理价值的微博意见领袖。Based on the nature of opinion leaders,the opinion mining technology is used to study the identification of microblog opinion leaders in emergencies in order to provide a reference for the governance of online public opinions.A three-step opinion leaders identification framework is proposed:first,using a literature analysis,an indicator model is built to evaluate the influence of microblog bloggers;second,a text subjective and objective classification model is built to determine whether high-influence bloggers publish event-related opinions and information,and to identify opinion blog posts;then an emotional polarity classification model is built based on the comment text of the opinion blog post to calculate the support degree of the blog post opinions and the support rate of the blogger’s opinions so that those high-influence bloggers who have first output their opinions and then received more support can be rated as the event"Opinion Leaders".These methods are used in the following text to analyze the microblog public opinion data of the typical practical cases,identifying the microblog opinion leaders of the incident,with their characteristics and public opinion participation behaviors observed and analyzed,and the results compared with those of social network analysis methods and the expert manual analysis method to verify the scientificity and effectiveness of this method.The results show that this method can identify microblog opinion leaders who truly have the guiding power of public opinions and can have governance value in the evolution of public opinions in emergencies.

关 键 词：观点挖掘 文本主客观分类 文本情感极性分析 突发事件 意见领袖 

分 类 号：TP3-05[自动化与计算机技术—计算机科学与技术] G202[文化科学—传播学]