Identification Of Malicious URL’s Using Machine Learning for Proactive Cyber Threat Prevention

Abstract

The rapid growth of internet usage has significantly increased cyber threats, especially through malicious URLs. These URLs are commonly used in phishing, malware distribution, and fraudulent activities. Traditional blacklist-based detection methods fail to identify newly generated malicious links. This project proposes a machine learning-based approach to identify malicious URLs proactively. The system extracts meaningful features from URLs and applies machine learning models to classify them as safe or malicious. Advanced algorithms such as XGBoost improve detection accuracy by learning complex URL patterns. Additionally, the system provides an evidence summary to explain why a URL is classified as malicious. This approach enhances user awareness and strengthens cybersecurity defences. Keywords: Malicious URLs, Machine Learning, Cyber Security, XG-Boost, Phishing Detection ,Feature Extraction, URL Classification