Analyzing and Recognizing Android Malware via Semantic-Based Malware Gene

Abstract

With the popularity of mobile platform, users' sensitive information and financial security are connecting closely to smartphones. An increasing number of malware samples appear in Android since it is the most popular operating system. However, there are no systematic methods to analyze and recognize those malware samples. Malware is labeled and classified in different standards. In this paper, Android malware gene is defined and extracted to recognize Android malware systematically. A malware gene is the minimum subsequence of statements to result in the functional information and it commonly occurs in a malware family. Moreover, a clustering via K-means is utilized to validate the identification effect of Android malware gene. Experimental results show that it is effective to recognize and analyze Android malware by means of malware gene.