Reverse Engineering Mobile Apps for Model Generation Using a Hybrid Approach

Abstract

The popularity of mobile devices is ever increasing which led to rapid increase in the development of mobile applications. Model-Based testing can improve the quality of mobile application but the models are not always available or are of inadequate quality. Reverse engineering approaches are used to automatically generate model from the GUI of mobile applications for model-based testing. This paper proposes a hybrid approach for reverse engineering mobile applications which exploit the capabilities of both static and dynamic approaches while trying to maximize the quality of the generated models. The insight of this approach is to use static analysis on app’s source to identify supported events. The generated events can be used to dynamically explore an app at run-time to generate a state model of the app’s GUI. The preliminary results from our approach indicated that the technique can generate high quality models from android apps.