Comprehensive URL Classification System using Machine Learning

The exponential growth of online content has necessitated effective systems for organizing, categorizing, and filtering web resources. URL classification, the process of assigning labels or categories to URLs based on their content, is fundamental in web content management, search engines, cybersecurity, and various other applications. This study is focused on analyzing and identifying a robust machine-learning model for classifying URLs into 4 categories: benign, defacement, phishing, and malware. A combination of datasets from 3 reliable sources along with a custom dataset has been used to ensure the reliability of the input data for the models. Six models were tested and compared using the F1-score. The Random Forest and Extra Trees Classifiers showed the best performance with accuracy of 98.81%.