Suspicious Link Detection using AI

Abstract

With the increase in internet usage, cybersecurity has become a major concern for computer systems, as malicious URLs can release different forms of malware and attempt to collect user data. The global lockdown in 2020 led to a significant rise in the use of internet services for business, which in turn resulted in a surge of cybercrimes committed by cybercriminals and significant data losses for businesses. To prevent such attacks, it is important to identify malicious URLs and understand the types of threats they pose. Signature-based approaches are often used to find such websites, and security tools are deployed to impose access restrictions on them. This chapter proposes using the linguistic aspects of related URLs to enhance the effectiveness of classifiers for identifying dangerous websites through Machine Learning algorithms such as Logistic Regression and Random Forest Technique. The study shows that being able to identify spam URLs solely based on URLs and categorizing them without relying on page content can lead to significant resource savings and a safer browsing experience for users.