Fighting Phishing the Information-Centric Way

Recent research efforts have highlighted the role of information as key target future network architectures. These so-called information centric networks base their operation on the information itself, rather than on other aspects or entities of the architecture, such as location, host, or user identity. In this paper we apply this approach at the application layer of the current Internet in order to mitigate a significant threat: phishing attacks. We designed a mechanism capable of detecting phishing Web pages based on the information that the user perceives while browsing, i.e, the mechanism detects attacks based on the browser´s graphical output, rather than on the code of the Web page. We implemented this mechanism in the form of a browser extension and we evaluated it using real data from a publicly available database of phishing URLs. Our extension, which is not affected by the location, popularity, or implementation of a Web page, detected 81% of the phishing attempts.