SCAM Detection in Credit Card Application

Abstract

Identity crime is well known, prevalent, and costly, and credit application scam is a specific case of identity crime. The existing no data mining recognition system of business rules and scorecards and known scam matching have confines. To address these confines and combat identity crime in real time, this paper proposes a new multilayered discovery system complemented with two additional layers: communal detection (CD) and spike detection (SD). CD finds real social relationships to reduce the suspicion score, and is tamper unaffected to synthetic social relationships. It is the whitelist-oriented methodology on a fixed set of attributes. SD finds spikes in false to increase the suspicion score, and is probe-unaffected for elements. It is the attribute-oriented approach on a variable-size set of elements. Together, CD and SD can detect more types of attacks, better account for changing legal activities, and remove the redundant elements. Experiments were carried out on CD and SD with several million real credit applications. Results on the data support the suggestion that successful credit application scam patterns are sudden and exhibit sharp spikes in false. Although this research is specific to credit application scam recognition, the concept of flexibility, together with adaptively and quality data discussed in the paper, are general to the model, implementation, and evaluation of all recognition systems.