Community-aware dynamic network embedding by using deep autoencoder

•This paper studies the embedding problem (DNE) in dynamic networks.•We design a deep autoencoder (CDNE) for solving the DNE problem.•We present a community-aware smoothing technique for DNEs.•CDNE preserves both the structure and evolution of dynamic networks.•We validate the superiority of CDNE over several state-of-the-art DNE methods.