Connections between the Sznajd model with general confidence rules and graph theory

Area of Science:

• Sociophysics
• Computational Social Science
• Network Science

Background:

• The Sznajd model simulates opinion propagation and consensus, favoring larger agreeing groups.
• Previous work generalized bounded confidence rules to model biases and prejudices in discrete opinion models.
• This study extends prior research by applying these modified rules to the Sznajd model.

Purpose of the Study:

• To link properties of mean-field fixed points in the Sznajd model to qualitative aspects of confidence rules (biases/prejudices).
• To establish connections between opinion dynamics and graph theory problems.
• To investigate the impact of group size on opinion formation dynamics.

Main Methods:

• Mathematical analysis of mean-field fixed points.
• Application of graph theory concepts: strongly connected graphs and maximal independent sets.
• Simulations on Barabási-Albert networks for comparison with mean-field results.

Main Results:

• Existence of fixed points is linked to strongly connected graphs.
• Stability of fixed points is connected to finding maximal independent sets.
• No qualitative difference in mean-field results when group size (q>2) is considered, aligning with the q-voter model.

Conclusions:

• The study provides a rigorous mathematical framework connecting opinion dynamics, biases, and graph theory.
• Mean-field analysis accurately predicts behavior in complex networks like Barabási-Albert networks.
• The findings offer insights into consensus formation and the role of social biases in opinion propagation.