ABSTRACT
This chapter analyses the flow of information within and between members of different communities, studying the dynamics of their interactions and the role of automated accounts in such exchanges, using Twitter as a benchmark. It focuses on the most effective accounts in tweets propagation by taking advantage of statistical physics techniques, and machine learning techniques to single out the automatic accounts operating on the network. The chapter demonstrates the impact of our approach by considering the propagation of Italian tweets concerned with two topics: migration flows in the Mediterranean and Covid-19 related discussion. It shows that bots play a central role in the exchange of significant content, and that the so-called hub nodes have, among their followers, a high number of bots. The activity of bots consists mostly of retweeting human users, although there is a non-negligible activity of genuine accounts in retweeting automated ones.
