Detection of phishing attacks in financial and e-banking websites using link and visual similarity relation

Today, phishing is one of the biggest problems faced by the cyber-world. In this paper, we present an approach that can detect phishing attacks in commercial and e-banking websites using the link and visual similarity relations. Phisher always tries to mimic the visual design of the webpage and the fake webpage contains identity keywords and hyperlinks that point to the corresponding legitimate webpage to trap internet users. Therefore, our proposed approach analyse the keywords, hyperlinks and CSS layout of the webpage to detect phishing attack. In the proposed approach, we make a set of associate domains with the suspicious webpage and explore the link and similarity relation to identifying phishing webpages. Also, we use the login form and whitelist based filtering to increase the running time of the proposed approach. Our proposed approach is not only able to detect phishing webpages accurately but its source webpage also. Moreover, it does not require any prior training to detect zero hour phishing attack. Experiments are conducted over a 6,616 phishing and legitimate webpages and the proposed approach gives approximately 99.72% true positive rate and less than 1.89% false negative rate.