Rumor Propagation And Control Strategy Based On Nonlinear Dynamics

With the development of Internet technology and the emergence of online social media,the ways of rumor propagation become more complex,the propagation speed is more rapid,and the range of influence become more wider,which cause more serious social harm.We are urgently required to understand the rule of rumor propagation,restrain the spread of rumor,thereby reducing the loss caused by rumor propagation.Therefore,it is of great practical significance to study rumor propagation model,reveal rumor propagation mechanisms,and propose effective control strategies to restrain rumor propagation.Until now,there are still some deficiencies in researching on the modeling and controlling of rumor propagation in social networks.In the aspects of modeling and analysis of rumor propagation,current studies are presented without considering the influence of the people who just discuss rumor topics and without using real datasets to verify the validity of the proposed model;while as for the aspects of controlling rumor propagation,most researches use a single control strategy but the cost of the control strategy is not taken into account.To address the above issues,this thesis aims to conduct research on modeling and controlling of rumor propagation.The specific research contents and innovations of this thesis are as follows.(1)Modeling and analysis of rumor propagation in social networks This thesis proposes a rumor propagation model that considers the influence of discussants,which divides the entire population into four categories:ignorances,discussants,spreaders and removers.By calculating equilibria and the basic regeneration number of the model,and analyzing the local and global asymptotic stability of equilibria and the transcritical bifurcation phenomenon,the rumor propagation mechanism is revealed.The correctness of the above theoretical analysis is verified by carrying out numerical simulation experiments.By comparing with the existing models,the results show that discussants have an important influence on the spread of rumor.Based on a real dataset,the least square fitting method is used to estimate all the parameters of the model,and the parameter values obtained by the fitting method can be used to predict the rumor propagation trend,thus the model can be verified.The simulation results show that the value of the goodness-of-fit R2 of the model is 0.9544,which indicates that the model can accurately simulate the rumor propagation in real social networks.(2)Research on rumor propagation control strategy.On a basis of the proposed model,a rumor hybrid control strategy that combines protection and blocking strategies is proposed,and the dynamic behaviors of the proposed strategy are analyzed.A real dataset is used to predict the propagation trend of rumor spreaders and truth spreaders in real social networks.The effectiveness of the hybrid control strategy is illustrated by comparing with the following cases namely without control strategy,a single protection strategy,and a single blocking strategy.In order to achieve best control of the rumor propagation with lowest cost,an optimal control scheme of the hybrid control strategy is proposed.Using the Pontryagin’s minimum principle,the existence of the optimal control is proved and the optimal control system is obtained.The experimental results illustrate that the cost of the optimal control is the lowest by comparing with the cost of the hybrid control strategy,the single protection strategy,and the single blocking strategy,and minimize the density of rumor spreaders.