A Study of Rumor Spreading with Epidemic Model Based on Network Topology

Abstract

The history of rumors might be as long as the history of human beings. Not much is known about this phenomenon until very recently when social networks provide a more efficient way for online rumor spreading. Micro-blogging platform, like Twitter and Sina Weibo, provides an ideal environment not only for rumor spreading, but also fruitful data to reveal the underlying rules controlling the birth, spread, and death of rumors. In this paper, we try to answer the following questions which have been confusing people for years. How do rumor propagate in the network? How do human factors affect the spreading patterns of rumor? How do we construct models to understand the collective group behavior based on probabilistic individual choices? Our study on micro-blogging systems help to expose the digital traces of online rumor spreading, which offers an opportunity to investigate how netizens interact with each other in the process of rumor spreading. Based on the analysis of the interactions, we propose a mathematical model running on real topologies to simulate this process and explore the characteristics of online rumor spreading. Experiments show that our model can reflect the transmission pattern of real rumors. Our work is different from previous ones by pointing out that a subset of the population can work as “Resisters” who are proactively persuading their neighbors not to believe the rumor, while previous works only label them as the “silent” nodes. To the best of our knowledge, this is the first time to emphasize the contribution of “Resisters” and to mathematically study their influence on the rumor spreading process.