Development of a Smart Mobile Application to Scan Vulnerabilities in Android Applications

Abstract

The rapid transition toward mobile-centric governance has established smartphones as the primary computing platform for essential public services, necessitating the processing of high-stakes citizen data. However, a critical disconnect persists between the generation of sophisticated technical vulnerability reports and the interpretability of these findings for non-technical stakeholders within government organisations. This study develops and evaluates MySecure, a smart vulnerability scanning and proactive forensic framework specifically designed for the Malaysian government mobile ecosystem. The research utilizes a hybrid methodological approach, integrating static analysis of application components with proactive forensic snapshotting to capture volatile digital evidence. Vulnerability prioritisation is operationalised through frequency-based Multi-Criteria Decision Making (MCDM) techniques, specifically TOPSIS and EDAS, to rank influential features such as dangerous permissions and exported components. Evaluated against a large-scale dataset and prominent Malaysian applications including MySejahtera and KWSP i-Akaun, the results demonstrate that frequency-based ranking of intents achieves a detection accuracy of 99.10%. Furthermore, the framework successfully reconstructed 100% of event timelines in reboot-free scenarios, identifying critical flaws such as cleartext HTTP traffic and weak cryptographic indicators. The study concludes that bridging the gap between technical auditing and user-centric risk explanation is essential for organizational security efficacy. MySecure advances the field of proactive forensics by operationalising automated standards into a functional, rootless tool tailored for high-stakes government architectures.