| In the era of Internet technology development,people are accustomed to communicating on the Internet.Large-scale users and intricate user relationships constitute a colorful online social network.Thousands of hot topics emerge and spread rapidly through various channels.Therefore,the social networking platform has become a paradise for all kinds of rumors.Due to the particularity of rumors and the limited number of users,the spread of rumors will be affected by anti-rumors and derivative rumors in social networks.How to effectively measure the impact of anti-rumors and derivative rumors on current rumors,as well as effective prediction of current rumor propagation trends,is an issue that needs to be studied and solved.This thesis mainly analyzes and studies rumor’s propagation in social networks from two perspectives,namely anti-rumors and derivative rumors.The main research work can be summarized as follows:1.In studying anti-rumor,we fully consider the relationship between rumors and anti-rumors,and construct a rumor propagation model based on evolutionary games and anti-rumors.Firstly,as for the symbiosis of rumors and anti-rumors,based on the SIR model of infectious disease dynamics,anti-rumors is introduced as an independent information into the spreading process of rumors to construct the SKIR model.Secondly,in view of the confrontational characteristics of rumors and anti-rumors,as well as people’s profitability and herd mentality,the Evolutionary Game Theory is adopted in this paper to construct the drive mechanism of user participation in the topic,and explores the causes of user behavior dynamics in the spreading process of rumors.At the same time,we combined internal factors and external factors of the users to build the influence of information by multivariate linear regression method,which provides the theoretical basis for user behavior driving force.Finally,we combined with the SKIR model proposed in this paper to model and analyze the propagation of rumors under the influence of anti-rumors.2.For derivative rumors,considering the competing relationship between current rumors and derivative rumors,this paper builds a rumor propagation model based on Lotka-Volterra model and derived rumors.First of all,due to the complexity of user behavior in the social network platform,the multiple linear regression method is used to construct the user’s behavior driving force from the user’s behavior and influence of neighbor node.Then,considering the competitiveness between derivative rumors and current rumors,and the dynamics of rumor influence,the Lotka-Volterra model is used to measure the competitive relationship between derived rumors and current rumors.Finally,we combine the traditional SIR model to construct the state transformation equation and model the propagation situation of the current rumors under the influence of derived rumors.In order to verify the validity and accuracy of the proposed model,this paper uses Sina Weibo,Tencent Weibo and network synthesis data for experimental verification.Experiments show that the model proposed in this paper can effectively reflect the influence of anti-rumors as well as derived rumors on current rumors,and can accurately predict the propaganda trend of rumors.This model helps in predicting the evolutionary trends and development laws of rumors in social networks,and at the same time provides a theoretical basis in internet rumors control as well as the dissemination of multiple messages. |
