Research On Attention Degree Of Weibo Events Based On Hidden Markov Model

With the rapid development and wide application of web 2.0 technology,self-media gradually entered the public's attention.In terms of social media,We Chat Moments,Qzone,and Weibo took the top three.Currently,the most popular self-media platforms in foreign countries are Facebook and Twitter,and Sina Weibo is based on "Twitter" as the most popular Weibo network platform in China.In March 2020,MAU(Monthly Active User)Reached 550 million and DAU(daily active users)reached 241 million.In recent years,Sina Weibo has exerted a significant influence on the proliferation of many real social events.This has become a double-edged sword for social development.It can not only promote the expression of public opinion and play a positive role,but also provide a way for the dissemination of false information.The hotbed has affected social security and stability to a certain extent.The tracking of Weibo event attention based on event classification management is an effective way to solve this double-edged sword.In order to predict the event attention more accurately,motivated by observations of social events' influence concerning with users and microblogs,we quantify the user popularity from the four dimensions of user activity,user behavior,user authenticity and user infection ability,the Pearson correlation coefficient between the four dimensions is calculated to test the non-collinearity between them.In turn,the comprehensiveness and non-redundancy of the evaluation results are ensured.Then,through the correlation analysis of user influence and attention of Weibo events,combined with HMM logic framework,we propose an algorithm to predict the Weibo event attention by using the user popularity.Meanwhile,in order to better detect the performance of the prediction algorithm,we integrate the static and dynamic information of microblog content to directly quantify the current Weibo event attention as a benchmark,and by calculating the similarity between the content of Weibo and its topic,the attention of Weibo events is corrected.Finally,based on the data sets of six real Weibo hot events that occurred in 2019-2020,the algorithm of this article(Hidden Markov Prediction Model HMM)and other three common prediction algorithms(Gray Prediction Model GM,Artificial Neural Network Prediction Model ANN,ARIMA Model)for comparative analysis.Through comparison,we find that the user popularity can be used to predict the event attention,and the HMM prediction method by using the user popularity shows good prediction performance.Three contributions are presented in this work:(1)we present a user popularity metric to measure user influence by combining comprehensively the characteristics of microblogs and users;(2)the Weibo event attention degree is defined on original microblogs and retweets;(3)Through the correlation analysis of user influence and attention of Weibo events,in the framework of HMM,the original observation sequence(user popularity) is used to predict the hidden state sequence(Weibo event attention).