View article

Fraud detection and prevention systems are based on various technological paradigms but the two prevailing approaches are rule-based reasoning and data mining. In this paper we claim that ontologies, an increasingly popular and widely accepted knowledge representation paradigm, can help both of these approaches be more efficient as far as fraud detection is concerned and we introduce a methodology for building domain specific fraud ontologies in the e-government domain. The main characteristic of this methodology is a generic fraud ontology that serves as a common ontological basis on which the various domain specific fraud ontologies can be built. The methodology along with the generic fraud ontology consist a powerful conceptual tool through which knowledge engineers can easily adapt ontology-based fraud detection systems to virtually any e-government domain.