Research On Information Communication Prediction Based On Micro Blog Opinion Leader

The rapid growth of social network stems from the booming of the Internet.Through social networking platforms,instant communication and News transmission have become one of the essential behaviors in people's daily life.The study of social networks is conducive to the accurate delivery of advertising products,the analysis of public opinion and the provision of scientific basis for social networks.In the research of information prediction for social network platforms at home and abroad,the influence of opinion leaders has the following problems: the selection of indicators is more subjective,and there is few researches combining real data when constructing influence;in the research of information dissemination prediction model The prediction accuracy of traditional models needs to be improved and the predictions are mostly displayed in real-time forwarding volume per unit time.There are few studies to predict the cumulative forwarding amount.Aiming at the above problems,an information propagation prediction model combining Bessel function and Ebers-Moll model is proposed.The main research content includes the following parts:First,we study the influence calculation method of opinion leaders,select the relevant indicators by fitting the real data crawled by Weibo,and calculate the weight of the indicators by AHP analytic method to calculate the influence of opinion leaders.Finally,opinion leaders' influence is graded depending on different scope.Secondly,we construct the information propagation prediction OLB model based on Weibo opinion leaders,combine the traditional model construction process,use Bessel function and Ebers-Moll model to find the correlation function relationship,and fit the opinion leader influence and prediction according to the selected indicators.The relationship between the amount of forwarding and the addition of factors such as the amplification series and time control parameters for the construction of the information propagation prediction model.Thirdly,we use the crawler to crawl the experimental data,select different levels of opinion leaders to compare the OLB model,the traditional model and the real data simulation experiment,and carry on the error analysis to judge the validity of the OLB model based on the errors.In summary,based on the characteristics of Weibo and opinion leaders,the information dissemination prediction OLB model is constructed,and the opinion leader influence calculation,information propagation prediction model establishment,data crawling and simulation experiment.Through simulation analysis and error analysis,the coincidence degree between the model and the real data is high,and the accuracy is improved compared with the traditional model,which indicates that the model constructed in this paper is reasonable and effective.