Exploring The Influence Of Structural Indicators On Information Dissemination In Micro-Blog System

Microblogging system offers a great business platform for viral marketingcampaigns. To maximize the information diffusion, many attempts are made to studyfeatures of information flow in user network. Previous studies mainly localized ineither the simple structural analysis of user’s network or information flow within thenetwork created by theoretical model, and hence, no reliable results can be providedfor microblogging system particularly. This paper studies the informationdissemination process based on both real user network data and informationdissemination path in microblogging system.Four phases are conducted in this study. The first phase focuses on literaturereview work;5indicators to describe the network structure are summarized;6indicators used to measure the information dissemination process are developed; thepotential influence of structural indicators on information dissemination indicators areproposed. The second phase collects26347nodes and998360lines as the usernetwork and150messages in microblogging are tracked. The third phase is toanalyze the network structure and compute the structural indicators by social networkanalysis. In the fourth phase, the influence of structural indicators on informationdissemination are studied and discussed.The main findings indicated the followings：(1) the degree-centrality performsexcellently in measuring information dissemination ability of an individual.Betweenness centrality performs excellently in measuring the control ability of anindividual. For closeness centrality, only if the value is higher than80%users in thenetwork, the indicator is capable to measure the coverage rate.(2) weak tie is criticalas a bridge in the information dissemination process. It controls many nodes andlevels along the dissemination path, however, strong tie is effective when passingthrough information within a cohesive sub-group.(3) the core is weak in controllingthe information dissemination process, though, it does exist in the network structurally.(4) By comparison of three centrality indicators, betweenness centralityis the most important indicator in the seed selection process.