基于社区发现的网络舆情主题—情感动态监测模型研究  

Research on Network Public Opinion Topic-emotion Dynamic Monitoring Model Based on Community Discovery

作　　者：江长斌[1] 王周阳 王虎[1] 李汾萦 Jiang Changbin;Wang Zhouyang;Wang Hu;Li Fenying(School of Management,Wuhan University of Technology,Wuhan Hubei 430070)

机构地区：[1]武汉理工大学管理学院,湖北武汉430070

出　　处：《情报探索》2023年第2期56-64,共9页Information Research

基　　金：国家社会科学基金项目“大数据视域下‘隐性’政治舆情的演化规律和治理路径研究”(项目编号:19BSH013)成果。

摘　　要：[目的/意义]探究热点事件网络社区的主题及情感波动,掌握舆情事件的演化过程,对于做好网络舆论引导、维护网络信息生态具有重要意义。[方法/过程]以社交关系构建社群网络,采用Louvain算法划分主题社区,基于此提取各阶段社区主题,并计算主题情感分布,多维度协同探究网络舆情的动态演化过程;最后以“孙海洋寻子圆梦”事件为例检验了模型的有效性。[结果/结论]网民会自发性地关注并讨论舆情事件的发展,并依据网络社交关系形成不同的主题社区,从而产生不同的情感波动。研究结果能够较好地反映出特定舆情事件的主题-情感演化过程。[Purpose/significance] It is of great significance to explore the themes and emotional fluctuations of online communities of hot events network communities and grasping the evolution of public opinion events for guiding online public opinion and maintaining the network information ecology. [Method/process] The paper builds a community network with social relations,uses the Louvain algorithm to divide topic communities,and then extracts community topics at each stage,and calculates topic emotion distribution,so as explore the dynamic evolution process of network public opinion in multiple dimensions;finally,“Sun Haiyang successfully found the kidnapped son” event is taken as an example to test the validity of the model. [Result/conclusion] Netizens will spontaneously pay attention to and discuss the development of public opinion events,and form different theme communities based on network social relations,resulting in different emotional fluctuations. The research results can better reflect the topic-emotion evolution process of specific public opinion events.

关 键 词：网络舆情 动态监测 社区发现 主题挖掘 情感波动 

分 类 号：G206[文化科学—传播学]