Abstract
- The digital transformation is unstoppable. The Internet of Things forms the basis of our digital information society. Growing cybercrime is one of the greatest challenges in this context. Targeted and complex cyberattacks on converged network infrastructures require innovative detection strategies to discover and classify multi-level malware and its underlying intentions. In this work, a modular framework is presented that enables reliable detection and classification of suspicious or malicious behaviour. The data basis for this is a preceding hybrid app analysis. The program consists of four modules. The pattern recognition and characterization of malware are based on a multi-stage supervised machine learning model, the functionality of which has been validated in initial system tests