Opinion Dynamics with Local Interactions (IJCAI-16)

Dimitris Fotakis, Dimitris Palyvos-Giannas and Stratis Skoulakis

This repository contains the code used in the simulations that are described in the paper. The full PDF is available here.

Abstract

We study convergence properties of opinion dynamics with local interactions and limited information exchange. We adopt a general model where the agents update their opinions in rounds to a weighted average of the opinions in their neighborhoods. For fixed neighborhoods, we present a simple randomized protocol that converges in expectation to the stable state of the Friedkin-Johnsen model. For opinion-dependent neighborhoods, we show that the Hegselmann-Krause model converges to a stable state if each agent’s neighborhood is restricted either to a subset of her acquaintances or to a small random subset of agents. Our experimental findings indicate that for a wide range of parameters, the convergence time and the number of opinion clusters of the neighborhood-restricted variants are comparable to those of the standard Hegselmann-Krause model.

Dependencies

The following python packages are required to simulate the models:

• numpy
• scipy
• networkx (network creation/display)
• tqdm (progress reporting) The following packages are needed if you want to use extra features (plots/parallel processing):
• matplotlib
• seaborn (can be ommited if you remove the relevant import statements)
• ipyparallel (for parallel processing)