AMD is a carefully-labeled and well-studied dataset that includes comprehensive profile information of malware.
AMD contains 24,553 samples, categorized in 135 varieties among 71 malware families ranging from 2010 to 2016. The dataset provides an up-to-date picture of the current landscape of Android malware, and is publicly shared with the community.
AMD provides detailed description of the malware’s behaviors through manual analysis. We produce detailed documentation of the manual analysis results and report our observations of the current landscape of Android malware.
Android malware apps are evolving towards monetization and more advanced anti-analysis techniques. Virus-writers are becoming more skilled and putting more effort to design comprehensive and sophisticated standalone malware.