Student research abstract: android malware detection based on Kullback-Leibler divergence

Abstract

A recent study shows that more than 50% of mobile devices running Google's Android mobile operating system (OS) have unpatched vulnerabilities, opening them up to malicious applications and malware attacks. The starting point of becoming a potential victim due to malware is to allow the installation of applications without knowing in advance the operations that an application can perform. In particular, many recent reports suggest that malware applications caused unwanted billing by sending SMS messages to premium numbers without the knowledge of the victim [1, 2]. Given that, there is a need for techniques to identify malicious behaviors of applications before installing them.