Model of opinion spreading and its dynamical studies in complex network

Abstract

In this century, many aspects involve information transmission in networking system. Not only in computer network, but also society and even biochemistry. Studying opinion spreading dynamics over different network models helps us to understand some phenomena about the world. In this project, a modified model is proposed based on an original rumor spreading model. The modified model simulates multiple opinions (e.g. opinion0, opinion1,..., opinionW) spreading on different network models. Three network models are random graph, small world and scale free network models. The opinion spreading model allows a node to accept or reject, or propagate, or forget opinions. And some opinions have larger chance to be accepted by a node (e.g. the chance of adopting opinion0> that of opinion1>...> that of opinionW). And the project focuses on how 1) network topologies and 2) opinion features affecting the dominance of opinions. Also, we study how 3) initial opinion spreaders affecting time to arrive opinion steady states. The results are as follows: For part 1, high average degree proposes the dominance of opinion0 for random graph and scale free network models. Small link rewiring probability in a small world network supports the dominance of opinionW. For part 2, opinion transmission rate affects the dominance of opinions strongly while the importance of opinion transmission probability is relatively small. For part 3, if initial opinion spreaders have higher centralities, the opinion steady time is usually shorter. Surprisingly, if there is suitable distance between those initial opinion spreaders, although they have a little bit lower centralities, opinion steady process sometimes is faster than that only depends on node centralities one.