Accurately labeling apps as malicious and benign is fundamental for training effective and reliable ML-based malware detection methods. The infeasibility of manually labeling apps forces researchers to rely on online platforms, such as VirusTotal, to label apps. Unfortunately, such platforms are often volatile and dynamic. The main objective of this thesis is to provide the research community with methods to optimally utilize VirusTotal scan reports until a more stable, reliable platform is implemented.
«
Accurately labeling apps as malicious and benign is fundamental for training effective and reliable ML-based malware detection methods. The infeasibility of manually labeling apps forces researchers to rely on online platforms, such as VirusTotal, to label apps. Unfortunately, such platforms are often volatile and dynamic. The main objective of this thesis is to provide the research community with methods to optimally utilize VirusTotal scan reports until a more stable, reliable platform is impl...
»
Translated abstract:
Da es sich letztlich nicht durchführen lässt, eine große Anzahl von bösartigen Apps manuell zu analysieren und zu kennzeichnen, um exakte Labels zu erhalten, sind Wissenschaftler dazu gezwungen, sich auf Onlineplattformen, wie VirusTotal zu verlassen. Vor diesem Hintergrund ist das Hauptanliegen dieser Arbeit anderen Wissenschaftlern Methoden zur Verfügung zu stellen, mit denen sie die Scan-Berichte von VirusTotal optimal nutzen können.