A Comparative Analysis of Machine Learning-Based Website Phishing Detection Using URL Information

The growing popularity of the internet help to expand e-commerce business, but such activities have security challenges caused by cybercriminals defrauding and stealing personal and financial information through website phishing. Phishing attacks are becoming increasingly complex and challenging to detect. An anti-phishing machine learning technique can help to identify a legitimate website from a phishing website by extracting various features from various sources like URL, page content, search engine, etc. This paper represents a comparative analysis of machine learning (ML) based website phishing detection and discusses some ML approaches for identifying phishing. We have compared five machine learning techniques-Decision tree, Random Forest, KNeighbors, Gaussian Naïve Byes, and XGBoost. Essential features which contribute much to the accuracy of the results have been chosen. The results show that the random forest algorithm outperforms the other proposed algorithms with an accuracy of 97.0 %.