Fraud matrix: A morphological and analysis-based classification and taxonomy of fraud

Abstract

A comprehensive taxonomy of fraud is presented based on morphological analysis, attribute listing and matrix analysis. Fraud matrix and tree classification frameworks are presented and discussed. They are then utilized to classify and explain a number of the different types of frauds, shown in the fraud matrix classification framework. First, triangular attributes of fraud are formulated, followed by fraud channels and elementary fraud features. Several well-known fraud types are identified using the proposed fraud classification framework. Further, new fraud types are discovered using the framework, for example, transactional frauds, automated frauds, synchronized fraud, unwitting accomplice, and ‘Robin Hood’ fraud. The importance of the taxonomy is that it can be used to classify both existing and newly identified fraud types in a way and manner that has not been previously reported; that is, it offers understanding to the different classes of frauds, the inherent threat actors behind such frauds, their capabilities, intent and the resulting nature of the frauds. This taxonomy has potential to offer insight in how appropriate countermeasures to mitigating the different types of frauds could be formulated.