Modeling Opinion Dynamics: Ranking Algorithms on Heterogeneous Populations

TL;DR

This paper presents an agent-based model of opinion dynamics that incorporates individual heterogeneity and social influence, enhanced with a ranking algorithm to reflect real-world social network behaviors, analyzed through a mean-field approximation.

Contribution

It introduces a novel agent-based model that accounts for demographic and structural heterogeneity, combined with a ranking algorithm and a mean-field analysis of its macroscopic behavior.

Findings

The model captures the influence of individual attributes on opinion formation.

The mean-field approximation provides insights into the system's macroscopic dynamics.

Properties of the autonomous system are thoroughly analyzed.

Abstract

The heterogeneity of the influence processes is an important feature of social systems: how we perceive social influence and how we influence other individuals is heavily influenced by our opinion and non-opinion attributes. The latter include demographic, cultural, and structural (how we are embedded in social networks) characteristics. Furthermore, the results of the influence processes may also depend on how similar the interacting individuals are in terms of their features. This paper addresses this issue and elaborates on an agent-based model that is sensitive to the individual characteristics, both opinion and non-opinion ones. The model is fortified with a ranking algorithm that mimics the ranking algorithms widely adopted in real-world online social networks. For the resulting model, I elaborate a mean-field approximation that describes the behavior of the model at the macroscopic level via an autonomous system of ordinary differential equations. The properties of this system are thoroughly studied.