Title:Detecting Cluster with Temporal Information in Sparse Dynamic Graph

Abstract:Detecting and keeping track of clusters in a dynamic graph is challenging, especially when all snapshots are too sparse to obtain any meaningful clusters. Instead of clustering at each snapshot, we propose a temporal clustering framework based on PARAFAC Decomposition to detect clusters and temporal information on how they change. We provide a deeper understanding of PARAFAC from clustering point of view. We apply our method to both synthetic and real-world data, compare the result with state of the art and show that our methods can detect clusters as well as their meaningful moment.