Font Size: a A A

Counter Rumor Models And Multi-rumors Spreading Dynamics On Complex Networks

Posted on:2016-05-14Degree:DoctorType:Dissertation
Country:ChinaCandidate:Y L ZanFull Text:PDF
GTID:1108330461984403Subject:Operational Research and Cybernetics
Abstract/Summary:Request the full-text of this thesis
The spreading of rumor and epidemic in the real-world social network are closely concerned with people’s daily life and economic activities. And the dynamics research about all kinds of spreading phenomenon on complex networks has been one of study hotspots in the complex system field for many years.In this paper, considering some practical phenomena about rumor spread-ing and studying the rumor spreading mechanism from several different re-search perspectives, we build several rumor spreading models and investigate the effects of some different types of influence factors in rumor spreading pro-cess both in theory and in simulation. The main innovation achievements are listed as follows:1. Two new rumor spreading models considering the existence of dis-senter are established, and the effect of self-resistance of complex networks to rumors is studied particularly.For the same rumor, different people may have different opinions. People would spread rumors which he agrees, while prevent those he disapproves. In chapter 2. considering the counterattack mechanism of the rumor spreading, this dissertation establishes two new rumor spreading models:Susceptible-Infective-Counterattack-Refractory (SICR) model and adjusted-SICR model, and then derives mean-field equations to describe their dynamics in homo-geneous networks and conduct the steady-state analysis. The dissertation also introduce the self-resistance parameter T, and study the influence of this parameter on rumor spreading. Numerical simulations are performed to compare the SICR model with the SIR model and the adjusted-SICR model, respectively, and we investigate the spreading peak of the rumor and the final size of the rumor with various parameters. Simulation results are con-gruent exactly with the theoretical analysis. The experiment reveals some interesting patterns of rumor spreading involved with counterattack force.2. Double-rumors concurrently spreading dynamics for two different ru-mors is researched, and a conclusion is gotten that the optimal launch time and the best promulgator can make new rumor do best to effect the old rumor.In the real social network, it is seldom that only one rumor is spread, but more of multiple rumors with various kinds of contents co-existing at the same time. Due to the above situation, this paper studies the spreading dynamics of multi-rumors co-existing on complex networks in chapter 3, and builds correspondingly two double-rumors spreading models:DSIR (Double-rumors concurrently spreading) model and C-DSIR model, respectively.Different from previous studies, the models here do not restrict the con-tents of two rumors, and also do not set rule in advance that one rumor has absolute predominance and prior to another, but provide that the two rumors would be selected by spreader in the spreading process. And then numerical simulations are performed to research the influence of the start-ing time of new rumor and the selection parameter. The results show that proper opportunity of launching a new rumor has to been found, and the best release time would obtain the optimal result to impact the transmission of old rumor. The closer to this best release time, the more apparent the influence to the spreading of old rumor. Besides, the more k-shell of the new rumor promulgator, the better to control the old rumor’s spreading.3. The role of the independence and impressionability of individuals in the rumor spreading process is studied, and the influence of individuals’ attitude and some interesting patterns of rumor spreading are discovered.In the real life, the individuals are under the influence of different fac-tors, which can be both opinions of themselves and affect from their friends. Meanwhile, the friends of a person have different importance or different level of friendship, and the more intimate of the friendship is, the more significant influence it is from the friend. So, it is necessary to take the personal opinion and the influence from neighbors into account when study the rumor spread-ing dynamics. And in chapter 4, this dissertation introduces a new rumor spreading rule from the dissertation of individual in the networks, then a list of simulations are performed about the two factors mentioned above. The research results show that the different confidence levels of an individual, to be more believe oneself or to be more affected by neighbors, has an obvious impact on the rumor spreading process, and the results reveals one another interesting patterns of rumor spreading involved in real life, the more out of pure friendship to share topics, the broader the topics with same interesting would be spread. While the more considering of the popular degree or status in selecting propagation objects, the worse the spreading result is.Above all, the research results about rumor spreading in several models is helpful in policy making to prevent and control the diffusion of rumors in real life, and the models established in this paper may have a certain approach value in studying the spreading mechanism and the spreading dynamics of rumors in real-world networks.
Keywords/Search Tags:complex network, rumor spreading, multiple rumors, pre-vention and control strategies, spreading threshold
Request the full-text of this thesis
Related items