Agent Based Rumor Spreading in a scale-free network (1805.05999v2)

Abstract: In the last years, the study of rumor spreading on social networks produced a lot of interest among the scientific community, expecially due to the role of social networks in the last political events. The goal of this work is to reproduce real-like diffusions of information and misinformation in a scale-free network using a multi-agent-based model. The data concerning the virtual spreading are easily obtainable, in particular the diffusion of information during the announcement for the discovery of the Higgs Boson on Twitter was recorded and investigated in detail. We made some assumptions on the micro behavior of our agents and registered the effects in a statistical analysis replying the real data diffusion. Then, we studied an hypotetical response to a misinformation diffusion adding debunking agents and trying to model a critic response from the agents using real data from a hoax regarding the Occupy Wall Street movement. After tuning our model to reproduce these results, we measured some network properties and proved the emergence of substantially separated structures like echochambers, independently from the network size scale, i.e. with one hundred, one thousand and ten thousand agents.