Statistical Properties of Portals' Users Activity

Abstract

We investigate the behavior of users at the web portal. We focus on transitions between two subpages, using the information about the path of the users. We construct a weighted network with portal subpages being its nodes and the weights of edges defined as a number of transition from between different subpages. We scrutinize the properties of such a network as a distribution of weights and strengths that corresponds to the power-law relation, as well as the statistics of users behavior. The latter brings interesting results: the average number of distinct subpages visited during the n-th step of a visit scales as a square root relation.