A grey-box approach for automated GUI-model generation of mobile applications

Abstract

As the mobile platform continues to pervade all aspects of human activity, and mobile applications, or mobile apps for short, on this platform tend to be faulty just like other types of software, there is a growing need for automated testing techniques for mobile apps. Modelbased testing is a popular and important testing approach that operates on a model of an app's behavior. However, such a model is often not available or of insufficient quality. To address this issue, we present a novel grey-box approach for automatically extracting a model of a given mobile app. In our approach, static analysis extracts the set of events supported by the Graphical User Interface (GUI) of the app. Then dynamic crawling reverse-engineers a model of the app, by systematically exercising these events on the running app. We also present a tool implementing this approach for the Android platform. Our empirical evaluation of this tool on several Android apps demonstrates that it can efficiently extract compact yet reasonably comprehensive models of high quality for such apps. © 2013 Springer-Verlag.