摘要
[目的/意义]在舆情领域,通过对已知主题生命周期演化轨迹的分析、热点与非热点主题演化过程的对比,可以更好地把握热点主题演化规律。[方法/过程]本文提出将TF-IDF算法融合评论影响力选择主题特征词,在此基础上,通过计算主题强度与相似度提出了6种主题演化形式,并在主题演化阶段将主题强度与主题内容两方面相结合进行了可视化实验。[结果/结论]最终展示出各个时间窗里的主题内容及主题强度,分析与挖掘出舆情事件中网民观点随时间的演化形式与演化规律。
[Purpose/Significance]In the field of public opinion,the evolution law of hot topics can be better understood by analyzing the evolution trajectory of the life cycle of known topics and comparing the evolution process of hot topics and non-hot topics.[Method/Process]In this paper,TF-IDF algorithm was proposed to select topic feature words by integrating comment influence.On this basis,six kinds of topic evolution forms are proposed by calculating topic strength and similarity.In the evolution stage of the theme,a visualization experiment was carried out by combining the strength of the theme with the content of the theme.[Result/Conclusion]Finally,the theme content and theme intensity in each time window were displayed,and the evolution mode and evolution law of netizens'opinions in public opinion events over time were analyzed and excavated.
作者
丁晟春
刘笑迎
李真
Ding Shengchun;Liu Xiaoying;Li Zhen(Department of Information Management,School of Economics and Management,Nanjing University ofScience and Technology,Nanjing 210094,China)
出处
《现代情报》
CSSCI
2021年第8期87-97,共11页
Journal of Modern Information
基金
江苏省社会科学基金项目“面向突发事件应急决策的知识服务智能化研究”(项目编号:20TOQ004)
江苏省研究生科研创新计划“基于主题发现的突发公共卫生事件舆情引导策略研究”(项目编号:SJCX20_0145)。
关键词
微博评论
评论影响力
网络舆情
主题演化
社会网络
TF-IDF算法
可视化
Weibo comment
influence of comment
network public opinion
theme evolution
social network
TF-IDF algorithm
visualization
