Opinion Dynamics on Correlated Subjects in Social Networks

TL;DR

This paper develops a mean field model to analyze how collective beliefs evolve on correlated subjects in social networks, considering individual dynamics, social interactions, and adversarial attitudes, providing insights into opinion stability and distribution.

Contribution

It introduces a multidimensional Fokker-Planck framework to model opinion dynamics with correlated subjects and adversarial interactions, advancing analytical understanding of social belief evolution.

Findings

Identifies conditions for opinion stability in social networks.

Derives steady-state opinion distributions based on personality and social factors.

Shows the impact of correlated subjects and adversarial attitudes on collective beliefs.

Abstract

Understanding the evolution of collective beliefs is of critical importance to get insights on the political trends as well as on social tastes and opinions. In particular, it is pivotal to develop analytical models that can predict the beliefs dynamics and capture the interdependence of opinions on different subjects. In this paper we tackle this issue also accounting for the individual endogenous process of opinion evolution, as well as repulsive interactions between individuals' opinions that may arise in the presence of an adversarial attitude of the individuals. Using a mean field approach, we characterize the time evolution of opinions of a large population of individuals through a multidimensional Fokker-Planck equation, and we identify the conditions under which stability holds. Finally, we derive the steady-state opinion distribution as a function of the individuals' personality and of the existing social interactions. Our numerical results show interesting dynamics in the collective beliefs of different social communities, and they highlight the effect of correlated subjects as well as of individuals with an adversarial attitude.