Machine Learning and Deep Learning for Phishing Page Detection

Abstract

The term "phishing" is often used to describe an attempt to obtain confidential data such as passwords or credit card details by impersonating a trustworthy source. In most cases, the term refers to attempts to trick users into providing sensitive information in response to a fraudulent email or web page. However, the term is also used to describe a broader category of online attacks to obtain sensitive information or to disrupt services or systems. Incorporating different machine learning and deep learning algorithms, including Support Vector Machine (SVM), Gradient Boosting Machine (GBM), and random forest, the authors of this research presented a technique for identifying phishing websites. The data sets from PhishTank and the University of New Brunswick were used to train and test the learning models. The XGboost model was able to surpass most existing techniques by achieving a maximum accuracy of 86.8%. This technique can be used in modern web browsers like Google Chrome and Mozilla Firefox to accurately detect phishing websites.