首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于关键词共现和社区发现的微博热点主题识别研究
引用本文:丁晟春,王楠,吴靓婵媛.基于关键词共现和社区发现的微博热点主题识别研究[J].现代情报,2018,38(3):10-18.
作者姓名:丁晟春  王楠  吴靓婵媛
作者单位:南京理工大学经济管理学院, 江苏 南京 210094
基金项目:国家社会科学基金项目"基于社会网络分析的网络舆情主题发现研究"(项目编号:15BTQ063);国家社会科学基金重大项目"面向知识创新服务的数据科学理论与方法研究"(项目编号:16ZDA224)。
摘    要:目的]从海量微博舆情信息中准确、高效地发现和挖掘当下的热点主题,以期为政府和企业监控和管理舆情动态提供有价值的参考。方法]研究首先从维度、特征和度量三方面综合考虑构建"帖子-主题"二模网络模型,其次选择词频、主题权重和词频增长率3个特征来抽取模型所需的有效关键词,然后选取社会网络分析法中的社区发现方法进行基于关键词的主题社区发现,最后综合考虑用户影响力和传播影响力两个属性,通过热度分析确立主题热度,识别热点主题。结果]实验表明,该方法能有效挖掘网络舆情中的热点主题,检测出的主题结果均正确,验证了本文方法的可行性和有效性。在此基础上,研究还选取实例进一步对主题的热度迁移和情感倾向进行了分析,具备一定的预警作用。

关 键 词:微博  热点主题  关键词  社会网络分析法  社区发现  热度迁徙  情感倾向  

Hot Topic Detection of Weibo Based on Keyword Co-occurrence and Community Discovery
Authors:Ding Shengchun  Wang Nan  Wu Jing
Institution:School of Economics and Management, Nanjing University of Science and Technology, Nanjing 210094, China
Abstract:Purpose] In order to find out and dig the hot topic from the weibo public opinion information accurately and efficiently,hoping to provide a valuable reference of monitoring and managing public opinion dynamic for the government and enterprises.Method] Firstly,this paper constructed the "post-topic" two-mode network model from the three aspects of dimension,feature and measure.Secondly,seleced the word frequency,word frequency growth rate and topic weight of the three characteristics to filter keywords needed by the model.Thirdly,it chose the community discovery algorithm in the social network analysis to do the topic discovery based on the keyword community.Finally,through the heat analysis to establish the topic of heat,it achieved the identification of hot topics.Result] Experiments showed that the method could effectively mine the hot topic in the network public opinion,the correctness of the detected topic results verified the feasibility and effectiveness of the proposed method.On this basis,the study also selected examples to further analyze the theme of the heat transfer and emotional inclination,with some early warning effect.
Keywords:weibo  hot topic  keyword  community discovery  heat migration  emotional inclination  
点击此处可从《现代情报》浏览原始摘要信息
点击此处可从《现代情报》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号