Analysis Method of Influence of Potential Edge on Information Diffusion

AbstractSocial phenomena are affected by the structure of a network consisting of personal relationships. In this paper, the diffusion of information among people is examined. Especially, we focus on the potential edge for two nodes that are not connected by an edge and have at least one common neighbor. First, a mechanism in which the potential edge changes into a real edge is considered and a new network model is proposed. This mechanism determines the topology of the network and the statistical indicators. Second, the role of a potential edge on the information diffusion is studied by numerical simulations using a simple information diffusion model of the networks. Two data mining methods are used: the neural network predicts the convergence rate and the time by six explanatory variables, and the decision tree reveals the statistical indicator having the strongest effect on the information diffusion. By analyzing the relationships between the information diffusion and the statistical indicators, the role of the potential edge is shown