Diffusion dynamics of competing information on networks

TL;DR

This paper introduces a microfounded threshold model to analyze how competing information spreads on social networks, revealing unpredictable outcomes and social polarization due to individual decision-making dynamics.

Contribution

It develops a novel threshold model incorporating individual optimization, providing new insights into the complex and indeterminate nature of competing information diffusion.

Findings

Virality of competing information is fundamentally indeterminate.

Diffusion follows a saddle path when individuals maximize coordination.

Irreversible choices lead to social polarization with multiple stable equilibria.

Abstract

Information diffusion on social networks has been described as a collective outcome of threshold behaviors in the framework of threshold models. However, since the existing models do not take into account individuals' optimization problem, it remains an open question what dynamics emerge in the diffusion process when individuals face multiple (and possibly incompatible) information. Here, we develop a microfounded general threshold model that enables us to analyze the collective dynamics of individual behavior in the propagation of multiple information. The analysis reveals that the virality of competing information is fundamentally indeterminate. When individuals maximize coordination with neighbors, the diffusion process is described as a saddle path, thereby leading to an unpredictable symmetry breaking. When individuals' choices are irreversible, there is a continuum of stable equilibria where a certain degree of social polarization takes place by chance.