Published Paper Details:

The role that the World Wide Web plays in enabling illegal actions including spam, fraud, and the distribution of malware is examined in this study three times.The focus is on malicious URL detection using machine learning, with Random Forest and MLP demonstrating high accuracy on a dataset of 2.4M URLs. The study addresses challenges posed by dynamic HTML development and proposes a resilient method for accurate malicious webpage detection in cybersecurity. Our approach, overcoming the limitations of traditional antivirus methods, analyzes webpage characteristics to identify malicious intent. In response to the escalating cyber threat landscape, we present a robust cybersecurity method that classifies websites based on URL features using supervised machine learning. The models, trained on a dataset of malicious and benign URLs, are evaluated with Random Forests, Gradient Boosted Decision Trees, and Deep Neural Networks. The paper also tackles challenges like the availability of training data to attackers and the dynamic nature of malicious websites, introducing a paradigm for detecting and countering induced concept drifts."