Detection of Malicious Websites Using Machine Learning

  • Dr.Kasiselvanathan.M, Sathiyapriya.A.M, Suruthimathi.C,Suryamitra.P


In Modern Internet world, security of private information is nightmare to every person due to the development of Web Apps, Mob Apps, etc., There are several ways for scammers to steal information from internet users, including web surfing, spam mail, and social media apps. The internet has become a forum for a wide variety of illicit activities, ranging from spam advertisements to financial fraud, thanks to technological advancements. Malicious websites contribute significantly to the growth of online illegal activity and stifle the advancement of Web services. As a result, there has been a significant push to create a systematic solution to prevent users from accessing such websites. Phishing is a malicious practice that involves inducing people to disclose personal details such as passwords and credit card numbers. It is one of the most common and least-protected security threats. Phishing attacks are dangerous threats that use human contact to persuade people to reveal sensitive information or take inappropriate acts. To recognize URLs, the proposed system only uses six features. The number of hyphens, dots, numeric characters, discrete variables that refer to the presence of an IP address in the URL, and the similarity index are all features of the URL. We suggest a learning-based method for categorizing Web sites into three categories: benign (safe), spam, and malicious. Our scheme achieves this by using machine learning algorithms such as the SVM Algorithm, MLP classifier, and Random Forest Algorithm.