Abstract:

Phishing detection is an area of identifying malicious activities designed by phishers to lure users providing sensitive information. Existing anti-phishing systems use blacklists based on specific parameters, characterize attacker’s activities with artificial and computational approaches and educate users. The development and maintenance of these systems is hard and costly because of the polymorphic nature of phishing techniques. Phishing attacks are able to scam humans with insufficient knowledge, while countermeasures focus on specific characteristics to make decisions. Defining formal approaches for representing and reasoning knowledge in anti-phishing systems is therefore a concern. This work deals with this issue by proposing formalized description logic to build the knowledge base of phishing attacks. It additionally designs an ontology-oriented approach to add semantics on that knowledge. The ontology model has been proven consistent and satisfiable. Experimentations on case studies demonstrate the ability of the proposed model to represent knowledge attack scenarios. A comparison with state-of-the-art researches shows that the proposed formalism is more adequate to characterize phishing semantics. This work could successfully complement anti-phishing systems.

Keywords: