BioND — Dynamics of Biological Networks

Opinion dynamics in social networks

Michael Seißinger
Bachelor thesis in physics, Julius Maximilian Universität Würzburg, 2010.


The dynamics of opinions is studied using directed voter models on adaptive networks. Analytical calculations applying the moment closure approximation are compared to data provided by simulations. Some properties of the analyzed directed models have already been discovered on undirected networks, such as a phase transition between global consensus and the splitting of the network into disconnected components with local consensus. Though, directed models provide a profound mathematical advantage, since the degree distribution is partially constant. Therefore, the directed voter models are useful in order to determine the influence of the degree distribution on the network's behavior. It is shown and explained that this influence is less important than it is predicted by the moment closure approximation.