The Impact of the Subgroup Structure on the Evolution of Networks: An Economic Model of Network Evolution

One of the most important properties of self-organized networks is their scale-free property. Prior research proved empirically and theoretically that scale-free networks emerge under the preferential attachment rule. However, a few empirical studies also show that empirical networks diverge from the structure of scale-free networks. Empirical networks exhibit a lower exponent of the power law distribution than constructed scale-free networks. Our research aims at establishing a simple evolutionary network model that explains this difference. The results of our model suggest that there are two reasons for this discrepancy. First, as already known, additional links between existing nodes distort the scale-free feature. Second, boundaries between subgroups (groups of network nodes) distort the degree distribution. In general, we believe that our evolutionary model may be applicable not only to describe the structural evolution of networks but also to make network design recommendations in a variety of areas such as WWW-hyperlink networks, business collaboration networks, Peer-To-Peer Networks, and Web2.0 service networks.