Discovery Of Tibetan Weibo User Groups Based On Topic Modeling And User Behavior Analysis

The rapid development of Internet web applications has driven the emergence of online social media platforms.Its rich user interaction features make more and more users accustomed to social activities on online social media platforms.In the social platform,users establish social relationships through various interactions,and as the interactions increase,the users constitute a community.With the increasing scale of Tibetan netizens,there are more and more Tibetan users in the online social platform represented by Weibo.The analysis of Tibetan microblog users in the Weibo social network has become particularly urgent and important.Community detection is a basic task in online social network analysis.It is only when a specific community is identified that the community can be further analyzed.For the community detection of Tibetan users in the microblog social platform,this article focuses on the following researches:1.Through analyzing the interaction behavior of users on the microblog social platform,this paper proposes a community detection algorithm driven by user behavior in a multidimensional directed network.This algorithm builds a user network structure based on the user's interaction behavior.Through the feature extraction and feature integration of the network structure to detect the community of specific user,experimental results show that the accuracy of the algorithm can reach 80.66%.2.For most Tibetan users using Chinese on social platforms,this paper proposes a community detection algorithm of Tibetan user based on user semantic information.The algorithm constructs user network structure by extracting user semantic information based on the improved LDA model,and then taking the similarity between users in the semantic information as the optimization goal to detect the community of Tibetan users in the microblogging network.The experimental results show that the accuracy of the algorithm can reach 81.09%.