PhishGuard: ML-Based Phishing URL Detector with API’s and Chrome Extension

In recent years, phishing attacks have been recognized as one of the most critical cybersecurity threats facing individuals, organizations, and financial institutions in the modern digital world. The attacks are based on exploiting human nature, wherein attackers trick users to access malicious websites, which are replicas of genuine websites, to steal confidential information such as login information, financial information, and personal information. However, it has been recognized that traditional methods such as blacklist-based and rulebased systems are no longer effective against modern phishing attacks, such as zero-day attacks, which are highly sophisticated and dynamic in nature. In this direction, this research proposes an innovative phishing URL detection system called PhishGuard, which utilizes machine learning to identify malicious phishing URLs. The proposed system is based on an extensive set of features, which include more than 40+ features (total 47 features), such as lexical, domain, and security-related features of URLs. The proposed system utilizes supervised machine learning models such as Random Forest, Logistic Regression, and Decision Trees to improve the accuracy of phishing URL detection. The proposed system also utilizes an external validation approach based on the API’s to improve its accuracy. The proposed system is light-weight, scalable, and can be deployed as a web application and browser extension. The proposed system has shown promising results in terms of improving phishing URL detection accuracy with minimal dependency on traditional static methods