The Sznajd dynamics on a directed clustered network

TL;DR

This paper studies the Sznajd opinion dynamics on a directed clustered network, revealing how network structure influences opinion persistence and convergence times.

Contribution

It introduces analysis of the Sznajd model on a directed clustered network, highlighting differences from fully connected networks in opinion dynamics.

Findings

Equilibrium time follows a log-normal distribution.

Equilibrium time increases linearly with system size.

Minority opinions can persist in the network.

Abstract

The Sznajd model is investigated in the directed Erdos--Renyi network with the clusterization coefficient enhanced to 0.3 by the method of Holme and Kim (Phys. Rev. E65 (2002) 026107). Within additional triangles, all six links are present. In this network, some nodes preserve the minority opinion. The time tau of getting equilibrium is found to follow the log-normal distribution and it increases linearly with the system size. Its dependence on the initial opinion distribution is different from the analytical results for fully connected networks.