Phishing Detection Using Machine Learning Overview and Significance

Phishing attacks have become a major cybersecurity threat, exploiting human vulnerabilities to steal sensitive information. Traditional detection methods often rely on rule-based systems and blacklists, which struggle to keep up with evolving phishing techniques. To address this, our project leverages deep learning models—Recurrent Neural Networks (RNNs) and Transformers—to develop a robust phishing detection system capable of identifying legitimate and malicious URLs with high accuracy. Our approach analyzes both static URL features (such as domain structure and character patterns) and dynamic contextual information extracted from website content. By combining sequential learning capabilities of RNNs with the attention mechanisms of Transformers, we aim to improve phishing detection beyond conventional techniques. The proposed system can assist cybersecurity frameworks in preventing phishing attacks and can be extended to detect phishing attempts in emails, text messages, and other online platforms.