An intelligent spyware detection system based on association mining

Abstract

The evolution and revolution of new engineering and technology requires stronger security mechanism to protect data and resources. Spyware is a general term, which is gradually becoming crucial for computer and network security resulting huge compromise in confidentiality, integrity and authenticity of the data source which impacts economy. In this research work, we provide the theoretical description of computer spyware and its types. We have proposed a definition of malware, which is explained in three cases. We have formulated a mathematical definition of malware propagation in a network using the basic concept of geometric progression; and also analyzed the infection rate in a simulated environment. The focus of this research work rate in a simulated environment. The focus of this research work is to detect spyware in user system. We proposed two spyware detection frameworks, for windows as well as for android. Windows spyware detection framework uses API calls extracted from different executable files whereas android spyware detection framework uses permissions from the applications. The dataset for the experiment is the combination of both primary secondary sources. We could gather a handsome amount of samples which includes malicious as well as benign applications for the experiment. FP Growth algorithm of association is used for the experiment. newline