基于交互行为的突发事件微博用户社区识别及研究

Identification and research of microblog user community in emergencies based on interactive behavior

针对现有研究对于微博社区舆情用户关系研究不足的情况,提出了基于微博用户交互行为的微博社区发现算法。首先通过突发事件下用户发布的相关微博内容对微博用户间的交互行为进行分析,将用户间交互行为进行量化,根据用户交互次数计算出用户交互度;其次将用户交互度作为Louvain算法权重对突发事件下微博用户群体进行社区识别;最后根据社区识别结果结合用户博文内容研究不同社区内用户关注主题。实验结果表明该方法能更好地对微博用户社区进行划分,且不同社区内用户的讨论与关注重点有所不同,因此在进行网络舆论治理时能更有效地引导网络舆论。

Aiming at the lack of research on the user relationship of microblog community in the existing research, this paper proposes a microblog community discovery algorithm based on the interaction behavior of microblog users. This paper firstly studies the interaction behavior among microblog users through the relevant microblog content published by users,quantifies the interaction behavior among users, and calculates the degree of user interaction according to the number of user interactions. Secondly, the user interaction degree is used as the weight of the Louvain algorithm to identify the community of microblog user groups, and finally, according to the results of community identification and the content of user blog posts, the topics of concern to users in different communities are studied. The experimental results show that this method can better divide the microblog user communities, and the discussion and focus of users in different communities are different. Therefore, it can guide network public opinion more effectively when doing network public opinion governance.