Community outlier detection in social networks based on graph matching

Abstract

Outlier or anomaly detection is widely used in several fields of study such as statistics, data mining, and social networks. It can reveal important anomalous and interesting outlier behaviours in the social network communities. In this work, we propose a new approach for community outlier detection based on social network graph matching. We identify community structures in social networks using some community detection methods. For each community, the node signature is combined with an optimal assignment method for matching the original graph data with the graph pattern data, in order to detect two formalised anomalies: anomalous nodes and anomalous edges. We also define a distance between two graphs using Euclidean formula. Then, we define a node-to-node cost in an assignment problem using the Hungarian method to deduce the matching function. The obtained experimental results demonstrate that our approach performs on real social network datasets.