Modeling And Simulation Of Rumor Propagation In Wechat Social Network

With the rapid increase in the number of netizen in China,more and more rumors spread in the WeChat social network.In order to better understand the rumor propagation mechanism and provide decision-making support for the management of WeChat rumor,the data of rumor in WeChat is systematically analyzed.Then a rumor propagation model is constructed.Finally,a simulation framework is proposed,with which the model is implemented,and the model is verified with the empirical data.The main research work and achievements of this paper are as follows:(1)Empirical analysis of rumor spreading in WeChat social network is conducted.There is no relevant result on the analysis of the rumor spreading data in WeChat social network for now,so the research results in this part fill in the blank in this research field.Our group collected a large amount of information data spreading in the WeChat social network,including the spread of rumor and non-rumor web pages.The user's viewing and sharing behavior as well as their corresponding timestamps and IP addresses are recorded in the dataset.The related characteristics of rumor spread in WeChat social network were obtained through data analysis.From the aspect of the topological structure,there was no significant differences between cascade size of rumor pages and non-rumor pages in WeChat social network.But there are significant differences between the two types of pages in the width of cascade,the height,the average depth and the average length of path.From the aspect of the temporal characteristics,rumors spread more rapidly and explosively than non-rumor pages.As for user behavior,we find that users respond faster to rumor pages than to non-rumor pages.(2)A WeChat social network rumor propagation model is constructed and validated.At present,many models cannot be validated because of the inability to find the right method to quantify and combine with the actual parameters.In the process of this research,a random Unknown-View-Share-Removed(UVSR)model is constructed which is a datadriven model.The related input parameters in the model come from the actual data,which solves the problem of no combination between theoretical analysis and actual data.The experimental results show that the cascade size,the cascade width and the average depth of simulation results of different network size have the same distribution trend with the actual rumor data.The result shows that the model can reproduce the related characteristics of rumor spreading in WeChat social network well.(3)An agent-based rumor spreading in social network simulation framework is proposed.This paper proposes an agent-based network rumor propagation simulation framework.The simulation framework divides the simulation process into three modules: the simulation environment configuration module,the simulation process execution module and the simulation result preservation module.The simulation environment configuration module can define the rules for different model,which reflects the universality of the simulation framework.The clear division of labor among each module is conducive to understanding the logic between modules and to enhance the readability of the model.In the experimental stage,the UVSR model was run on the simulation framework,and the feasibility of the model and the model framework was verified.This paper combines the theoretical model and the actual data in the research of rumors spreading in WeChat social network,and validates the model through experiments.The model has strong scalability.In the future research work,the model can be refined by refining the parameters of each part.