International Journal of Engineering and Technology

Phishing is a kind of attack that belongs to social engineering and this attack seeks to trick people in order to let them reveal their confidential information. Several methods are introduced to detect phishing websites by using different types of features. Unfortunately, these techniques implemented for specific attack vector such as detecting phishing emails which make implementing wide scope detection system crucial demand. URLs analysis proved to be a strong method to detect malicious attacks by previous researches. This technique uses various URL features such as host information, lexical, and other type of features. In this paper, we present wide scope and lightweight phishing detection system using lexical features only. The proposed classifier provides accuracy of 93% with 0.12 second processing time per URL.