Abstract

Android Malware has emerged as a consequence of the increasing popularity ofsmartphones and tablets. While most previous work focuses on inherentcharacteristics of Android apps to detect malware, this study analyses indirectfeatures and meta-data to identify patterns in malware applications. Ourexperiments show that: (1) the permissions used by an application offer onlymoderate performance results; (2) other features publicly available at AndroidMarkets are more relevant in detecting malware, such as the applicationdeveloper and certificate issuer, and (3) compact and efficient classifiers canbe constructed for the early detection of malware applications prior to codeinspection or sandboxing.

 

Quick Read (beta)