An efficient data enrichment scheme for fraud detection using social network analysis

With the continuous fast paste of ecommerce growth and the ever increasing amount of electronic transactions, we witness a similar if not more significant increase in electronic fraud. Consequently, fraud detection systems need to update their methods and improve them by considering more sources of information to stay effective. In this paper, a data enrichment scheme is proposed which concentrates on using social network analysis to help the detection system by feeding information that is hidden in the relations among entities. Since one of the challenges of a real life electronic transaction system is the large amount of data and number of users, the proposed scheme is presenting an efficient method to update the social network, as well. Simulation results indicate that the proposed scheme is able to detect fraud scenarios that are not detected using typical anomaly detection methods based on the normal behavior of cardholders. Hence, providing a higher accuracy, while minimizing the updating procedure.