As a new emerging social media, micro-blogging is becoming more and more popular among network users with its convenient interaction style and real-time platform feature. Users can freely post tweets on the platform, follow their interested users and quickly browse online information. The appearance of trending topics enriches users’ micro-blogging life. By attending trending topics that they are interested in, users can quickly acquire the desirable information, including interested users community, the evolution of these events and their favourite tweets. As data in the micro-blogging platform increases explosively, users’ demand for valuable information becomes more intense. To make full use of information elements on the platform, study on effective community discovery and community evolution analysis methods can provide a better analysis for user interest and alleviate information overload problem, which in turn provides high quality information services to micro-blogging users. The real-time convenient features of the micro-blogging platform provide users a natural approach to post tweets that are related to their interest, follow their interested people and acquire the information their followees have posted or re-tweeted. Users’ post behavior and the friendship network reflect their interest from different angles, by combining these two aspects information and mining the latent association between them, the community discovery research can help users quickly locate their favorite community types, which will provide users a new approach to browse valuable content information; The content that users within the same community are discussing and concerning changes over time. There is often a certain degree of connection around the related topics between user's interest community structures from different time periods, such as derivatives, disappear, splitting and merging. On the basis of user community mining work, researches on user communities’ evolution with topics can help understand the background of specific topics, analyze and model community users’ behaviors fully, which provides a more comprehensive analysis method for information organization and user information services. User communities will evolve with the passage of time, the specific phenomenon is that users’ behavior will be influenced by their friends’ earlier performance, which prompts community structure’s evolution with time. In the context of rapid inform...
修改评论