The number of followings as an influential factor in rumor spreading

In this paper, we hypothesize that the number of followings users have on social networks, has an influence on the impact level of the received rumor/anti-rumor messages by those users. We show that influence as a decreasing exponential coefficient on the probabilities by which users' might get affected from the received messages. Then we put forward another hypothesis that suggests that rumor/anti-rumor posts gradually lose their impact power by passage of time. To assign this time-related feature to the posts, we consider a memory for each agent of the network. Finally, we derive stochastic equations of the new model which incorporate both of the new hypotheses and then evaluate it based on real datasets of rumor spreading on Twitter. The evaluation results support the new hypotheses and show that the novel model is able to better represent rumor spreading on online social networks. (c) 2019 Elsevier Inc. All rights reserved.