The Empirical Analysis Of The Information Propagation In Online Social Network

Weibo has become an indispensable platform for people's public life.This personal media is playing an increasing influence and impact in the society,and its function is becoming more and more prominent.Therefore,it is of great value to study the influence mechanism of weibo and the diffusion mechanism of weibo to monitor the communication and evolution of microblog public opinion.This article mainly on sina weibo as the research object,and dissemination of information data mining and finishing the user data,and discusses the microblog information dissemination mechanism,weibo heat evaluation mechanism as well as the microblogging network information diffusion prediction and simulation,the main work can be divided into the following three points:（1）To the country's largest microblogging platform sina weibo as object,based on weibo attention mechanism,develop the relationship between the user data mining methods and rules,and according to the characters of weibo users individual development and optimize the weibo WAP end data acquisition platform,an excuse for developers to make use of sina weibo data mining work properly,the microblog content information,user information,weibo transmission of information and user information,obtain the relevant relationship between sina weibo user data and user data,and based on the user attention and weibo information dissemination relations according to obtain the real data to construct user relationship network and information transmission network.With classic SIR virus propagation model SEIR rumor propagation model,and the team for the improvement of SEIAR rumor spreading model,and according to the structure characteristics and weibo user network news spread store-and-forward mechanism combined with users put forward severe spread will be divided into four kinds of nodes:healthy,normal communication nodes,immune,severe communication node SI₁I₂R model.The simulation results show that the model is more consistent with the characteristics of microblog network information transmission.（2）According to the mining of user data and the participation of users for weibo process were analyzed,with the characteristics of the data to sort to find several factors affecting the heat of weibo network based on the data from sina weibo,including analyzing the characteristic of the information transmission in the microblog weibo blog end distribution,weibo propagation time and twitter spread participation,the propagation characteristics of various factors of influence the spread of weibo users based on weibo,including in characteristics and information dissemination eleven indexes such as,put forward the user's"heavy participation"and"negative participation"to participate in the weibo dissemination and the influence of heat,heat evaluation model based on factor analysis method was improved,and a single weibo measurement and statistics of various data characteristics,using the factor analysis sorts through the data analysis sorting,and finally the results and analysis with the original order difference causes,the results show that the model has high accuracy,this model can be suitable for weibo heat given parameters and sorting.（3）According to the interactive behavior of micro-blog users,the factors that affect the spread of micro-blog information spread are explored,and a prediction model suitable for the information diffusion of the micro-blog network event is proposed,and the path of the information diffusion of the micro-blog network event is described by the micro-blog user network and the micro-blog transmission network.Combined with Logistic mathematical model,we analyze the way of event information diffusion in micro-blog network,and complete the prediction of micro-blog network event information diffusion.The network simulation results show that the recall rate of micro-blog network event information diffusion prediction is high,and the prediction results are consistent with the actual results.