Trajectories of Evolving Communities in Social Networks

Abstract

This work attempts to provide a visual framework to track the evolution of communities in dynamic social networks. We resort to three-order tensors to represent evolving communities and we further decompose them using a Tucker3 model. The results of the decomposition are explored in order to obtain temporal trajectories of communities, by projecting them in a bidimensional space spanned by the components yielded by the Tucker decomposition.