URL Classification

URL classification aims to automatically categorize URLs as benign or malicious (e.g., phishing, spam, malware), a crucial task for cybersecurity and online safety. Current research focuses on improving the accuracy and robustness of classification models, employing techniques like convolutional neural networks (CNNs), graph neural networks (GNNs), and large language models (LLMs), often incorporating contrastive learning and adversarial training to enhance performance against sophisticated evasion techniques. The development of explainable models and efficient methods for handling large-scale datasets are also key areas of investigation, with the ultimate goal of providing more reliable and understandable URL security solutions.