A Novel Framework For Malicious URL Detection Using Hybrid Model

Abstract

Web searching is a process where a software program searches the database and gathers information related to the specified terms. Search is done with the help of search engines like Google, Bing, etc. Generally, the search engines offer search engines through which users are allowed to search for the required content via World Wide Web (WWW). The WWW includes massive amounts of data sites where it gets difficult to obtain the relevant data. These days, web users experience problems with information overload and sink because the amount of information and number of users has grown significantly and rapidly. In general, clients can enter keywords in the search engines to get the appropriate data. The search engines return a list of URLs as a result in a ranked order. Clients will most often choose the first link as the most relevant link for the requested data. However, sometimes it so happens that users may click the top-ranked URLs, which may not be the legitimate URL, and by doing so, users' data will be stolen by third-party attackers. Many scientific studies show that the number of methods is based on machine learning to detect malicious URLs. In this paper, we have proposed a hybrid model to see the given URL is a phishing URL or not.