Risks to network security of information are increasing both in quantity and severity at this time. The present-day hackers usually take advantage of end-to-end technologies and people’s vulnerabilities. These techniques include, among other things, phishing, social engineering, and pharming. One of the aspects of these assaults is tricking people with fraudulent URLs. Therefore, identifying fraudulent URLs is crucial right now. Numerous scholarly articles have addressed the topic of machine-learning-based fraudulent URL verification. This study presents a machine-learning approach that uses our suggested URL patterns and features to identify fraudulent URLs. In this case, the URLs are classified using the Random Forest and Multinomial Naive Bayes classifiers; the associated precision for classification rates are 97.19% for Multinomial Naïve Bayes, 97% for Random Forest, and 97.33% for Decision Tree.