Font Size: a A A

A Quantitative Study On The Influence Of Microblogging

Posted on:2015-08-16Degree:MasterType:Thesis
Country:ChinaCandidate:X P YiFull Text:PDF
GTID:2208330422467780Subject:Applied Statistics
Abstract/Summary:PDF Full Text Request
With the continuous development of social networks, microblogging hasbecome in the moment a lot of people interaction platform. Microblogging, asexplained in the Baidu Encyclopedia is a relationship based on user informationsharing, dissemination and access platform, users can set up individual communitiesthrough WEB, WAP and other clients to about140words of text updates, and instantsharing. The rise of microblogging has changed the traditional mass media, aprofound impact on people’s lifestyles and forms of social interaction. Based on someof the characteristics of its own microblogging, microblogging users sought after bymany people, which is a big influence. For microblogging influence extends in thefield, attracting a large number of researchers to carry out research, the paper alsocarried out research work.Firstly, the essential characteristics of the microblogging, and for the impact ofthe current problem microblogging platform zombie powder brought under thetraditional pagerank algorithm model and not an effective solution, to this endproposes an improved pagerank algorithm, the model reached a filter off zombiepowder aim to bring microblogging users of scientific and rational influence rankings.Secondly, its own characteristics for improved pagerank model andmicroblogging, found improved pagerank model there are still many deficiencies,the algorithm is too single-sided, therefore, proposed to user behavior-andcomments forwarded another factor, by Tony Yates forwarding probability theory toquantify, which made the correction model forwarding probability based model isalso better to avoid the influence of extreme values of the case, more reasonable toevaluate the overall impact of microblogging.Finally, the model experiments and results analysis, through detailed analysisand comparison to verify the improved pagerank model can effectively reduce theimpact of microblogging zombie powder as well as the influence of the model based on Bayesian probability of having a comprehensive and objectivity.
Keywords/Search Tags:microblogging, pagerank algorithm, zombie powder, Bayesianprobability
PDF Full Text Request
Related items