Abstract

Phishing attacks pose a significant threat to online security, exploiting unsuspecting users through deceptive websites that mimic legitimate platforms. This project presents a heuristic-based approach for detecting phishing websites using different ML techniques. The system is developed in Python and utilizes a robust dataset consisting of 11,055 URLs with 30 key characteristics to determine their legitimacy. Various machine learning models were evaluated, with LR demonstrating optimal performance, achieving a training accuracy of 95.00% and a validation accuracy of 92.00%. The Proposed system provides real-time analysis of URLs and enhances security by identifying Phishing attempts across multiple online platforms such as emails, SMS, social media, and websites. The project also includes an email spam detection module, further Strengthening protection against cyber threats. The developed model offers an efficient, accurate, and scalable solution to mitigate phishing attacks, contributing to a safer online environment.