Section: New Results

Exposure to diffusion in dynamic networks

In many contexts, complex networks are subject to diffusion phenomenon, like spreading of epidemics in human groups or the diffusion of information in social networks. Often, the underlying network is dynamic, that is, his links change along time. Clearly, the dynamics of links has an influence on the diffusion phenomenon taking place over the network. A first step to understand these relationship is to determine which nodes of the dynamic network are more likely to be reached by a diffusion process. We designed new notions of exposure in order to do it, based either on contacts, paths or flows in a dynamic network. In particular, the notion of dynamic flow, which we introduced, has given interesting preliminary results. We computed the exposition scores of nodes of real world dynamic networks and showed that it is correlated to the likelihood of nodes to be affected by a diffusion in the classical SI model.