A Novel Machine Learning based Hybrid Model for Webshell Detection

Abstract

Webshell attack has become a greater cause of concern while major episodes are shifting online. Today different forms of webshell attacks and attack inducing tools are available to hamper the security of computer systems. These attacks strongly escalate the requisite for Machine Learning based detection. In this work, we are going to obtain behavioral-pattern that may be achieved through static or dynamic analysis, afterward we can apply dissimilar ML techniques to identify whether it's web shell or not. Behavioral based Detection methods will be discussed to take advantage from ML algorithms so as to frame social-based web shell recognition and classification model.