This research addresses the growing threat of phishing attacks in the digital era by proposing an automated AI-based phishing website detection system. The project focuses on utilizing machine learning to distinguish between phishing and legitimate websites with high precision. Key highlights of this work include: Feature Extraction: The system leverages URL-based and domain-based features for effective classification. Model Evaluation: Three supervised learning models—Logistic Regression, Support Vector Machine (SVM), and Random Forest—were rigorously evaluated. Top Performance: Experimental analysis shows that the Random Forest classifier outperformed others, achieving a peak accuracy of 97.8%. Practical Use: The proposed system is designed for high efficiency and real-time applicability within modern cybersecurity infrastructures.
Taufikuzzaman Md. (Thu,) studied this question.