A dissipative network model with neighboring activation

Abstract

We propose a network model with dissipative structure taking into consideration the effect of neighboring activation and individual dissipation. Nodes may feel tired of interactions with new nodes step by step, and drop out of the network evolution. However, these dormant nodes can become active again following neighbors. During the whole evolution only active nodes have opportunities to receive new links. We analyze user behavior of a real Internet forum, and the statistical characteristics of this forum are analogous to our model. Under the influence of motivation and dissipation, the degree distribution of our network model decays as a power law with a diversity of tunable power exponents. Furthermore, the network has high clustering, small average path length and positive assortativity coefficients. Show more