A Proactive Approach for Fake Website Detection Using Machine Learning

Phishing is a cyber-attack technique that employs fraudulent websites to trick individuals into revealing sensitive information, such as passwords, usernames, and financial details. With the rising prevalence of phishing attacks, developing robust detection systems is essential to safeguard individuals and organizations. The proposed system leverages a dataset comprising features derived from website components, including URL structure, domain attributes, and content properties. In the preprocessing stage, the dataset is refined by cleaning the data and selecting pertinent features to enhance the identification of phishing websites. The system employs advanced classification models, namely XGBoost, CatBoost, and LightGBM, to accurately distinguish between legitimate and phishing websites. These gradient-boosting-based approaches enhance online security and mitigate the risks posed by phishing attacks. Keywords: Phishing, Cybersecurity, Machine Learning, XGBoost, CatBoost, LightGBM