Web Content Based Features for Malicious Web Page Detection Using Machine Learning

Malicious web pages are well-known threats in cyberspace. Malicious webpages and their URLs appear to be legitimate webpages in cyberspace and are used to steal personal information from internet users (victims) or to install malicious software on the victim's machine in order to gain further access and make the victims more vulnerable in cyberspace. Identifying such web pages in cyberspace is a difficult task for internet users. Researchers use different techniques with different sets of features to detect malicious web pages in cyberspace. Most of them extract vital features from URL text for malicious webpage classification due to their simplicity and risk-free nature. Some of them extract the limited number of features from the webpage content for better classification along with other features. Our proposed system considers features that are extracted only from the webpage content and use a machine learning model for classification. The accuracy of the existing and proposed systems is demonstrated experimentally. The result shows that the proposed system is more efficient.