A Sketch Based Game Theoretic Approach to Detect Anomalous Dense Sub-Communities in Large Data Streams

Abstract: Detecting anomalous subgraphs in a dynamic graph in an online or streaming fashion is an important requirement in industrial settings for intrusion detection or denial of service attacks. While only detecting anomalousness in the system by edge frequencies is an optimal approach, many latent information can get unnoticed in the process, since as a characteristic of the network only edge frequencies are considered. We propose a game theoretic approach whereby using the modularity function we try to estimate in a streaming graph \emph{whether addition of a new edge in the current time tick results in a dense subgraph creation, thus indicating possible anomalous score}. Our contributions are as follows: (a) We propose a novel game-theoretic framework for detecting dense subcommunities in an online streaming environment; (b) We detect such subcommunities using constant memory storage. Our results are corroborated with empirical evaluation on real datasets.