Phishing URL classification using Extra-Tree and DNN

Abstract

Machine Learning (ML) and Deep Learning (DL) methods have become indispensable in cybersecurity. Recently, they are often used to detect and classify phishing websites. Phishing websites are a major problem that has a negative impact on organization and of societies. Statistics report that the number of phishing website is continuously increasing and it is becoming more difficult to detect them. Various works have shown that ML and DL can be efficient to solve this problem. In this work, we adopted lexical analysis and Tiny URL approaches for URL features extraction. The accuracy metric obtained surpasses 98% for Extra Tree algorithm and can achieve 99% for Deep Neural Network model.