An XGBoost classifier machine learning model paired with a chromium browser extension that identifies phishing websites using features extracted from the URL. Also includes a adaptive learning component that keeps training the model based on user feedback. The goal of the project was to fight zero day phishing attacks and minimize the required human intervention for it.
The machine learning model is better than database matching, rule based, pattern based or any other static methods that are not effective against zero day attacks.
The model has a testing accuracy of 92%
Made with ❤️
