• 2020-11-10
• Ikuomola A.J, Ogunbameru A., Nwanze M.N
• Volume 2, Issue 2

Abstract

Malicious websites host unsolicited content and lure unsuspecting users to become victims of scams. It is imperative to detect and act on such threats in a timely manner. Traditionally, this detection was done mostly through the usage of blacklists features. However, blacklists lack the ability to detect newly generated malicious Uniform Resource Locators (URL). To improve the generality of malicious URL detectors, blacklist and machine learning techniques using feature extraction were explored in this design. Blacklist Feature takes lesser processing time and also relies on external data (list containing malicious websites) in detecting malicious websites while feature extraction method takes more time and does not rely on external data in detecting new malicious websites using the web browser extension. The system was implemented using JavaScript and MySQL. Certain malicious and benign websites were used to test run the system. The system consists of three major layers: Users, Web browser extension and Database. The web browser extension layer makes use of two techniques (Blacklist feature and feature extraction) to detect malicious website efficiently. The performance of the malicious websites detection system using both blacklist and feature extraction shows that it provides robust, secured and easier way to detect malicious website in real time.

Download PDF