Abstract
As mobile apps become more ubiquitous, there is an increased likelihood of more sophisticated cyber-crime (ransomware, spyware and zero-day) adversary threats among victims using traditional signature-based antivirus applications, which cannot keep pace with the stealth threat of the malicious software. The proposed architecture is to develop a multitier analysis model with a Artificial Intelligence (AI) Engine based on Machine Learning (ML) and Natural Language Processing (NLP) that develops classification set to user roles and threats previously identified from previous contexts to detect and block mobile malware in a real-time framework. The real-time framework assuages the user through improved real-time performance output with the behavioral analysis augmenting the static analysis process. The mobile malware detection model relies on knowledge gained from the application metadata through the NLP. The framework was designed for Android and provides the user realtime alerts warnings, isolation and an efficient means of analysis of threat material for the most at risk applications in order to conduct low footprint early malware detection for the user.