中文微博热点话题挖掘研究  

Research on Chinese Micro-Blogging Hot Topic Mining

作　　者：何跃[1] 帅马恋[1] 冯韵[1] 

机构地区：[1]四川大学商学院,四川成都610064

出　　处：《统计与信息论坛》2014年第6期86-90,共5页Journal of Statistics and Information

摘　　要：微博热点话题代表公众对现实生活问题的态度,对微博热点话题的识别有益于网络舆情监控。基于话题检测与跟踪技术设计出中文微博热点话题识别流程。首先通过K-means文本聚类归纳出话题,然后进行话题影响力计算和分析,最后通过话题影响力大小识别热点话题。实证结果表明,热点话题的"召回率"较高,影响力较大。话题影响力的构建为相关企业或政府针对话题热度大小采取不同的舆情监测策略提供了理论依据。Micro-blogging hot topic represents the public attitude to the problems of real life ,and the recognition of the micro-blogging hot topic is beneficial to monitor of network public sentiment .The paper based on Topic Detection and Tracking （Topic Detection and Tracking ,TDT） designs the recognition path of Chinese micro-blogging hot topic ,which analysis extracted micro-blogging hot topic by impact of topic to identify the hot topic .The result shows that according to the size of the impact of hot topics of the experimental period ,the ranking list is confirmed well by the evaluation of TDT . The topic influence building provides a theoretical basis for relevant enterprises or government to take a different control strategy of public opinion for the topic of heat size .

关 键 词：微博 话题影响力 话题检测与跟踪 

分 类 号：G203[文化科学—传播学]