Intelligent Agents in Securing Internet

Abstract

Internet phishing has become one of the most prevalent problems of online security. Phishing takes advantage of the user’s trust and use social engineering techniques to deceive them. As browsers have become a major platform for attacks to take place, a greater focus should be given for securing web services in this perspective.
This paper discusses a framework to resist phishing attacks utilizing the power of intelligent agents. The main focus is on a browser based attack called tabnabbing which executes in inactive browser tabs. The proposed method uses agents in three levels to monitor browser tabs, at regular intervals and warn the user at the earliest. This approach excellently protects user against Tabnabbing, URL obfuscations and malicious links. Results show that the proposed method outperforms the state of the art phishing detection methods and achieves an accuracy of 97.3%.