Research And Application On Microblogging Pollution Propagation Model

With the changing trend of Internet users from passive recipient to active creators, Web2.0has brought another revolution to Internet. The social networks develop rapidly with owning their large user groups. Among these social networks, microblogging leads the distance between Internet users to be further shortened. Accompanied with the increasing in its influence, microblogging becomes the juncture of guarding network’s security.Microbloggers can access to each other’s microblogging messages directly by their concern objects. Messages in microblogging are largely dependent on the user’s personal networks to spread. When a microblogging massage reaches the user, the next behavior is decided by the interaction between the activity of the source, the influence of receptor and the intimacy of the two users. Additionally, when it appears pollution in the microblogging network, how to locate the infected users in the large user groups quickly and efficiently, how to minimize the harm caused to the network by handled properly are the important researches need to be concerned.The main contents of this thesis are the analysis and simulation of rumor propagation characteristics in microblogging network, and this article proposes application named rumor detection model of microblogging, which is based on the microblogging information network. The study of influence user behaviors impact on the users’ relationship and the whole microblogging network is the most important part.This thesis starts from the analysis on characteristics of the three major components in microblogging:users, relations and messages, appropriate indicators are selected to quantify the interaction between the users and the message propagation model of microblogging, which is build based on the analysis, proposes the propagation coefficient impacted by both the sending and the receiving sides, distinguishes the interactive habits of different types of users with analytic hierarchy process, and builds the rumor propagation model in microblogging. By contrast with the actual rumors pollution observation experiments, this thesis verifies the rationality of the model. Then this article simulates different types of users on the model to study the influence on the rumor propagation impacted by the type of users. At the same time, the process of rumor propagation in microblogging is analyzed in the opposite direction, this article searches the previous source from the node the pollution is found, based on rumor propagation model of microblogging and ant colony algorithm, and keep expanding the scope of the search to locate the other infected users. By parameters discussions, the accuracy of search is improved, and the cost of cleaning up the contaminated network environment is saved with a great extent.