Modeling and quantifying the influence of rumor and counter-rumor on information propagation dynamics

Today, the spread of Internet rumors has become increasingly easier, however, once they spread widely, the network environment and even the social order will be broken down. Therefore, to help in designing effective rumor control strategies, we propose the two-stage rumor propagation dynamics model, in which the susceptible/educated-infected-recovered (SO-S/EIR) dynamics model is used to characterize the first stage and the susceptible/educated-infected-denied-recovered (C-S/EIDR) dynamics model is to characterize the second stage. In our model, according to the difference of rumor spreaders' confidence in the rumor, we divide them into super-infected and normal-infected, moreover, the global model incorporates vigilantes and rumor debunkers to reflect the diversity of public opinions. Our model is based on the forwarded quantity in the Chinese Sina-microblog and takes into account the impact of interactions between forwarding users, own cognitions, and own knowledge level on the propagation of the rumor. Our Model-based analytic and numerical analyses show that forwarding and transfer probabilities in our model are in accordance with the actual accident situation, besides, our sensitivity analyses describe important factors that affect the spread of the rumor in support of decision strategies to refute the rumor.