Malicious URL Detection with Explainable Machine Learning Techniques
Bolun Wang · 2025
Malicious Uniform Resource Locator (URL) detection research has been actively worked on and the focus is on achieving higher and more accurate results with machine learning and artificial intelligence approaches. Unfortunately, existing research on this problem has not yet been completed completely, and there remains a research gap in malicious URL detection in developing effective approaches that are capable of dealing with evasive techniques used by the attackers to conceal harmful URLs. This study explores an effective technique of detecting malicious URL detection with machine learnings with explainability. In particular, three advanced ML models are applied on one real parameters URL dataset, Logistic regression (LR), decision trees (DT) and Random Forest (RF) are employed. The results show that RF has the best detection accuracy and the best performance in terms of explain ability.