Arizona State University
When
Where
Title: Opinion formation and graph dynamics: from modeling to empirical applications
Abstract: The rise of social media has brought a paradox: even though people are more connected, opinions are becoming more segregated due to, for instance, the echo chamber effect. In this talk, we aim to discuss various models of opinion formation and their link to their interaction graph (i.e., adjacency matrix).
In the first part, we revisit a classical model of opinions (the Krause-Hegelsman model) and investigate how to find sufficient conditions for the emergence of consensus. The challenge is to consider influence functions with compact support (local interaction).
In the second part, we introduce a novel model with repulsion interactions, which also maintains opinions in a bounded domain. Rather than being repulsed, opinions could be attracted toward "opposite" opinions. We call this Doppelganger dynamics (each opinion has a double). Early investigations show how polarization and chaotic dynamics could emerge from these dynamics.