Phishing URL Detection

Abstract

Introduction

This project develops a machine learning-based system to detect phishing attacks by analyzing URL structures, domain features, and page content, offering an adaptive alternative to traditional blacklist methods. The model continuously learns from new phishing tactics such as domain spoofing and hidden redirects, providing real-time, automated protection with high accuracy.

Key objectives include collecting a quality dataset, extracting and optimizing relevant features, training an effective machine learning model (e.g., Gradient Boosting Classifier), and deploying it for real-time phishing URL detection.

Results show the system achieves 80% accuracy with quick response times and features like single URL prediction, PDF report generation, and a real-time monitoring dashboard. This makes it suitable for cybersecurity teams, corporate IT, and small businesses.

Future enhancements plan to include real-time alerts, mobile and cloud deployment, multilingual support, and adaptive learning to keep up with evolving phishing threats.

Conclusion

This project develops a machine learning solution to combat phishing attacks, where criminals create deceptive websites to steal login credentials and financial data. Unlike traditional detection methods that rely on outdated blacklists, our system analyzes URL structures, domain characteristics, and page content patterns to identify sophisticated phishing attempts. The model continuously learns from new threats, adapting to evolving attacker tactics like domain spoofing and hidden redirects. By automating detection with high accuracy, we reduce dependence on error-prone human verification

Copyright

Copyright © 2025 L Lavanya , Mythri Y, Rachana R, Rekha R. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.