Machine Learning-Based Detection of Phishing Websites Using URL, Domain, and Webpage Features

This research presents a machine learning-based approach for detecting phishing websites using 30 features derived from URL, domain, and webpage characteristics. The study evaluates four supervised learning models: Logistic Regression, Decision Tree, Random Forest, and Support Vector Machine on the UCI Phishing Websites Dataset containing 11,055 samples. Experimental results show that the Random Forest classifier achieves the best performance with 97% accuracy and an AUC score of 0.99. The proposed system is efficient, lightweight, and suitable for real-time phishing detection.