Abstract: Phishing is one of the most widespread and damaging cyber threats, exploiting human psychology and technological weaknesses to steal sensitive information. This project focuses on detecting and preventing phishing attacks by classifying various tactics used by attackers, such as fake websites, fraudulent login portals, and social engineering techniques. A machine learning–based detection engine is developed, supported by feature extraction from URLs, email headers, and web content. Different models including Random Forest, SVM, and Decision Trees are evaluated using accuracy, precision, recall, and F1-score metrics. The proposed system integrates a real-time browser extension and dashboard for monitoring phishing attempts, while also emphasizing user training and awareness. Experimental results demonstrate the effectiveness of combining technical detection methods with behaveoral education, thereby enhancing user protection. This research highlights the importance of adaptive, multi-layered approaches for combating phishing and contributes to building more resilient cybersecurity frameworks.

Downloads: | DOI: 10.17148/IARJSET.2025.12924