Published Paper Details:

Phishing websites have emerged as one of the most prevalent cyber threats, deceiving users into revealing sensitive credentials and causing significant financial losses. Detecting these malicious websites automatically and accurately is crucial for online security. This paper presents a machine learning-based phishing website detection system that analyzes multiple data modalities including URL lexical features, HTML structure, and WHOIS information to identify phishing attempts. The system extracts key features such as domain age, URL length, special character patterns, SSL certificate validity, and embedded script behaviors, which are then processed using ensemble learning models such as Random Forest and XG Boost for classification. The proposed model achieves an overall accuracy of 96.4%, precision of 95.8%, and F1score of 96.1% on benchmark datasets, demonstrating superior performance compared to existing methods. The model is lightweight and optimized for real-time deployment, making it suitable for browser extensions and online verification systems. The key contributions include the integration of diverse feature sets, enhanced detection accuracy, and a scalable implementation for real-world phishing prevention.