Understanding Information Diffusion And User Influence On Campus Network

With the development of Internet technology and a wide range of popular networkapplications, Internet has become the most important place of information sharing anddiffusion, computer based Internet interactions behavior has become an important mode ofuser communication and information dissemination channels.Especially for the majority ofcollege students, the campus network has become an indispensable part of life andwork.However, the essential characteristics of the Internet information diffusion andexchange have not formed a clear and profound understanding.Individual users in theinformation dissemination process of the role of analysis and can not accurately measure.Inthis context, the study of behavioral characteristics of the campus network users, andanalysis on the basis of the measured data through the network of informationdissemination process and user behavior information sharing between a cross-disciplinaryresearch focus.In this paper, information dissemination structure to the campus network,research network topology for the information dissemination and its regularity.Mainly forthe current main campus network in new media network communication channels: forums,social networking and micro-blog the three most typical online social networking platformfor network infrastructure, and information dissemination process of measurement andanalysis.Research internet user behavior, construct model of the network users impact,proposed algorithm to value user impaction, measure the importance of user in informationdissemination process. The research of the paper can help people to understand the networkinfrastructure of information diffusion process and basic characteristic and the complexhuman group behavior has important implications.The main thesis research and innovation in the following areas:1. We measure and analysis the network structure of the campus BBS, proposed aweighted network model to construct the user relationship in the bbs network. Throughin-depth analysis of the measured data, the BBS network has the character of scale-free andsmall world, showing that the network self-organization, and a small amount of core usercontrol of the network nodes of speech community.tell us the causes of the existence of theopinion leaders.Then we proposed a algorithm to assess the degree of the user authoritynamed LeaderRank, which is based analysis of emotional replies of the user. By thecompare the main opinion leader recognition algorithms, we found that the networkarchitecture can effectively help the identification of th opinion leaders. At last we analysis the relationship of the core user and the BBS broad and found that the most active useralways on a small number of BBS broad.2. We research the websit of renren as an example,analysis of the social networkinformation dissemination mechanism, the informaion diddemination process is divided totwo process,from outside to inside, and campus communication process between usersandestablishment of the corresponding cross-linked network user model and effectivepropagation path modell.The biggest innovation of the model is that it can be clearly seenfrom the microscopic point of information dissemination of information to infiltrate thecampus, and how the information is spread within the campus.Statistics found bymeasuring the relative concentration and there is not a fixed penetration of information tothe campus user community of external users, but the dissemination of information tocampus outside the first point of contact is a very small part of the crowd, so long as thecontrol of a small number of core internal users, external information is difficult to spreadon campus.So we proposed a spread of the influence of the user evaluation algorithm,considering the spread of individual will and social communication capacities of these twofactors, and through comparison such social indicators the amount of user visit, valid theaccuracy of the algorithm.3. By the analysis of the mechanisms of the sina microblog informaion dissemination,we build two informaition-sharing network models: user listening network model and usersharing network model. The innovation of the user sharing network model is theinformation diffusion process based on real user behavior. We got the actual data of thecollege access the sina microblog. Based of the data statistics and analysis, we found thatcampus users have a very strong characteristic of aggregation and willing to shareinformation. In terms of information dissemination, a small part of the user account for thevast majority of information and communication behavior.Then an algorithm is alsoproposed based on the pagerank named WeiboRank algorithm can real assessment thespread of the influence of the user. The results show that algorithm presented in this paperWeiboRank accurate dissemination of information on the user’s ability to assess whether theinformation in the user listening network or sharing network as a whole are significantlybetter than the comparison algorithms.