Malevolent WEBSITE identification using Machine Learning Techniques

Abstract

Threats to network information security are currently increasing both in number and seriousness. Today's hackers are mostly focused on attacking end-to-end systems and finding weak points in people. These strategies include pharming, phishing, and social engineering. These assaults include a stage where malicious Uniform Resource Locators (URLs) are used to trick visitors. Consequently, harmful URL detection is growing in popularity. The use of machine learning and deep learning algorithms to identify malicious URLs has been proven in several academic articles. In this study, we propose a malicious URL detection system that makes use of machine learning approaches based on proposed URL behaviors and attributes. Bigdata technology is additionally being utilized to enhance the detection of risky URLs based on peculiar behavior. A novel set of URL traits and behaviors, a machine learning algorithm, and big data technologies make up the suggested detection technique. The trials' findings imply that the suggested URL characteristics and behavior can significantly improve the ability to identify potentially harmful URLs. This shows that the suggested method might be regarded as the best and most convenient way to find counterfeit URLs.