Social Networks for Fraud Detection - University of Nottingham Malaysia Campus

Fraud is as old as humankind and appears in many types and forms. Popular examples are credit card fraud, tax evasion, insurance fraud, anti-money laundering, and payment transaction fraud. In earlier research we defined fraud as an uncommon, well-considered, imperceptibly concealed, time-evolving, and carefully organized crime. Nowadays, fraud is typically tackled using state-of-the-art predictive or descriptive analytical techniques with many accompanying challenges. In this talk, we elaborate on how social networks can be leveraged to detect fraud. We extensively discuss how to define social networks and how to carefully featurize them to boost the performance of analytical fraud models. Throughout the presentation, I will extensively report on both my academic and business experience and illustrate the concepts discussed with various real-life examples.