It is commonly perceived that the design principles of the future Internet might be drastically different from today. It is natural to ask what will be the impact of such evolution on the design of future Online Social Networks (OSNs). There is evidence that human social networks may be invariant with respect to the underlying online technology supporting them. Furthermore, the increasing pervasiveness of communication technologies is likely to enable any two users to communicate anytime and anywhere. Thus, a possible evolution of OSN design could map directly the structure of human social networks, and build future OSN services on top of a network whose edges represent "communication channels" between users sharing social relationships, and activated when they interact because of their social ties. In this paper we look, in the perspective of future OSN designed according to this concept, at how the patterns of interactions between people in human social networks impact on information dissemination properties. Based on well-established theories from the anthropology field, we study the properties of inter-contact times between users, i.e. the time between successive communication opportunities. This is a crucial feature for information dissemination, as previous results obtained in a conceptually similar environment have shown that the distribution of inter-contact times determines the convergence properties of information diffusion protocols. In the paper we investigate, by analysis, simulation and experimental results, the impact of different users interaction patterns on the properties of inter-contact times and, thus, on the convergence properties of information dissemination protocols.
展开▼
