The Android platform has grown beyond smartphones to support many device types, development frameworks, and architectural patterns. This evolution challenges traditional static analyzers designed for Java-based mobile apps.
The research identifies three key limitations in current approaches: device diversity, framework heterogeneity, and architectural modularity. The study created a dataset of phone-TV applications to examine device-specific security implications, developed a cross-language analysis methodology for React Native, and created a model for analyzing dynamically composed applications.
Their approaches significantly improved static analysis precision and completeness for modern Android applications, establishing a foundation for more adaptable analysis techniques in the future.