Abstract: Studying the dynamics of face-to-face interaction networks is essential for a better understanding of contact mediated phenomena, e.g. contagion processes, and disease spreading. Based on high-resolution proximity sensor data, recent studies analyzed the properties of time-dependent contact networks, revealing that these typically consist of a disconnected ensemble of dense groups and show strong circadian modulations. We devise a class of Markovian dynamic network models that naturally yield group formation and are easy to control with a small number of natural parameters. Introducing circadially varying activity rates, we show that face-to-face contact networks may be interpreted to follow a trajectory of equilibrium states in a newly found state space. Within this state space, we study how the epidemic response curve and the epidemic threshold of an SIS process change with varying group sizes and mixing time of the system, revealing non-trivial dependencies. It is found that even infinitely infectious diseases may die out under certain conditions.
Alongside the presented research a new C++/Python-package for the analysis and visualization of temporal networks will be demonstrated (github.com/benmaier/tacoma).