Identifying The Popular Microblog Based On LDA And Information Entropy

In recent years,along with the development of mobile Internet,short video,social networks,and other related areas are also growing.In the social network,with the development of foreign social network tools such as Twitter,Facebook,Instagram,the domestic social network began to rise.In our country,from the initial Fanfou,Renren,Blog,to the present Microblog,WeChat,the development of social networks continue to evolve.The original Microblog products include Sina Microblog,Tencent Microblog,Sohu Microblog.However,Sina Microblog stand in top of the microblog product list due to a huge user volume and opened a new social era.In this article,"Microblog "refers to Sina Microblog.Academic research for microblog is also deepening.Topic discovery,public opinion analysis,influence evaluation are now more popular research direction.A large number of Internet users have been accustomed to share their work,life,access to the latest information,hot spots through microblogging.Users of Microblog can publish text,pictures,video,and also forward and like the moment of other users.Users can review followed users'status on their own home,you can also browse other users' home page.Based on its huge user groups,microblog data updated and increase constantly.Information overload is a major problem for users currently.They need to spend a lot of time and effort to view the hot events related microblog.Because not all microblog information is important or valuable,found hot microblog from the massive microblog is an important and challenging task.Based on the above background,identifying the popular microblog is being put forward.The main research methods are as follows:First,the LDA(Latent Dirichlet Allocation)model based on information entropy is proposed to extract the microblog theme.The model can reduce the degree of uncertainty in the subject,so that the microblogging theme is more accurate.Second,calculating the internal heat and external heat of microblogging.Finally,combined with the ranking algorithm to structure microblog score function.This article take TOP300 microblogs as a popular microblog.The model can effectively find the popular microblog,the F-value of the model is higher than the existing research method,and more accurate.