Android Datasets: Integration of VirusTotal and Quark-Engine Intelligence

The datasets shared here are derived from an existing dataset (doi.org/10.34894/H0YJFT) containing 870,515 applications with 48 metadata features. We enriched this original dataset using external malware intelligence from VirusTotal Premium v3 Enterprise and Quark-Engine 25.3.1. Of the 870,515 original apps: - 803,320 apps (92.3%) were matched to VirusTotal reports. - 772,732 apps (88.8%) were successfully scanned by Quark-Engine. - 772,533 apps (88.7%) had paired analyses from both sources.

Based on this new intelligence, we generated new datasets that include parsed versions of the VirusTotal and Quark-Engine reports. Specifically, we enriched the original dataset with 10 new features (5 from Quark-Engine reports and 5 from VirusTotal). We also revalidated the original class label, producing 5 new validated datasets based on different VirusTotal detection thresholds.