Phishing Detection: Comparison between Uniform Resource Locator and Content-Based Detection

Authors: Nuur Ezaini Akmar Ismail, Norbazilah Rahim, Norul Huda Md Rasdi, Maslina Daud

Abstract:

A web application is the most targeted by the attacker because the web application is accessible by the end users. It has become more advantageous to the attacker since not all the end users aware of what kind of sensitive data already leaked by them through the Internet especially via social network in shake on ‘sharing’. The attacker can use this information such as personal details, a favourite of artists, a favourite of actors or actress, music, politics, and medical records to customize phishing attack thus trick the user to click on malware-laced attachments. The Phishing attack is one of the most popular attacks for social engineering technique against web applications. There are several methods to detect phishing websites such as Blacklist/Whitelist based detection, heuristic-based, and visual similarity-based detection. This paper illustrated a comparison between the heuristic-based technique using features of a uniform resource locator (URL) and visual similarity-based detection techniques that compares the content of a suspected phishing page with the legitimate one in order to detect new phishing sites based on the paper reviewed from the past few years. The comparison focuses on three indicators which are false positive and negative, accuracy of the method, and time consumed to detect phishing website.

Keywords: heuristic-based technique, phishing detection, social engineering and visual similarity-based technique