A Novel Approachfor Phishing Websites Detection using Decision Tree

Abstract

Phishing is a fraudulent activity conducted by attackers to deceive users and obtain crucial information. Phishers mostly use social engineering method to achieve their malicious goals. Website phishing has been a troublesome affair for organizations, common people and firms since years. Even SSL/TLS protected websites also fall in the trap of phishing attack. In this paper, we have also considered HTTPs websites and used machine learning method, specifically supervised learning for the detection of phishing websites. We have used J48 algorithm due to its proficiency and great outcomes. The classifier efficiently detects websites affected from phishing attacks on the basis of features selected fromphishing websites and genuine websites. The approach provided in this paper manages to successfully predict legitimate and phishing websites, offering 95% accuracy rate. Confusion matrix is generated to depict False negative rate, True negative rate, False positive rate and True positive rate respectively.