Agent Modeling and Simulation of Rumor Propagation Behavior Evolution in Social Network

Abstract

The spread of rumors in social networks shows greater concealment, subjectivity and aggression, is tend to be measured in the form of exponential functions. Based on the traditional model of infectious diseases, the social network rumor propagation model is investigated by differential equation. Then the simulation model via Anylogic is proposed by defining the agent of Internet users (USER), rumor maker (RM), and rumor breaker (RB) firstly. Then, the states of susceptible, exposed, infectious, mongering and recovered are defined in USER agent, at the same time the properties of transitions are set. A function of rumor_breakFUN to generates "rumor" message was defined. The rumor_breakFUN function is just to be executed at the beginning of experiment which indicates rumor maker immediately exit from social media platform after distributing rumor message. Then an influence function of the rumor break is defined in rumor breaker agent, which could be used to judge which Internet user is covered by the effect variable attach to the Internet user is set to true. The user’s state conversion probability is changed when rumor breaker works on the user. At last, the changes among the five different states of the Internet users in the main agent are represented with different colors, and the changes of the number of users are shown in the time plot. The work theory analyzes the rumor propagation mechanism in the context of social networks. The simulation model reveals the transformation rate and number of social media users between different states under the condition of specific state transition parameters. It shows the influence mechanism of rumor dispeller in the process of rumor spreading. Through the analysis of the simulation model, the importance of the rumor breaking in time is discussed, and some suggestions on the problem of rumor management are proposed.