Research On Influence Of Microblog User Based On HQRank Algorithm

In recent years,with the rapid development of the Internet,network applications have become an important channel for people to access information,especially the network social platform-microblogging.In this thesis,we mainly take Sina as research platform and focus on the problem of user influence analysis.The conclusions in this study could be useful for public opinion monitoring and leaders mining in microblogging.Three major contributions of this paper are as follows:(1)Microblogging user influence research.Based on the analysis of user behavior and information dissemination,researching a variety of data indicators for users in the microblogging.The user's influence is defined as the user and content of microblogging impact on public and contribution to information dissemination.(2)User influence evaluation algorithm HQRank.In this paper,the PageRank algorithm,Influence-Rank are deeply interpreted and analyzed.The advantages and disadvantages are summarized.Considering the first two algorithms in defects of the evaluation of the individual importance of the user,fans influence and microblogging influence are put forward.The improved h index calculate the quality of fans to avoid zombie powder.Microblogging quality is calculated by the number of transponders and comments.The influence of fans and microblogging influence as a basis for the weight matrix of the PageRank algorithm.The new HQRank algorithm is created for microblogging user influence ranking.(3)Achieve the HQRank algorithm with Hadoop framework.In this paper,The HQRank algorithm is implemented step by step with the convenience of Hadoop's distributed computing.Finally,capturing the user data on Sina microblogging.HQRank algorithm is verified by experiments.The result of experiments show that HQRank algorithm ranking is more credible than Influence-Rank algorithm in terms of user influence.