EvoCreeper: Automated Black-Box Model Generation for Smart TV Applications
This addresses the need for developers to model and evaluate smart TV apps, particularly for quality assessment in contexts like smart houses, but it is incremental as it applies existing black-box methods to a new domain.
The paper tackles the problem of analyzing smart TV applications by presenting an automated black-box reverse engineering strategy to generate models of app user interfaces, which can be used to assess app quality and completeness, with evaluation showing it automatically generates unique states and comprehensive models.
Smart TVs are coming to dominate the television market. This accompanied by an increase in the use of smart TV applications (apps). Due to the increasing demand, developers need modeling techniques to analyze these apps and assess their comprehensiveness, completeness, and quality. In this paper, we present an automated strategy for generating models of smart TV apps based on black-box reverse engineering. The strategy can be used to cumulatively construct a model for a given app by exploring the user interface in a manner consistent with the use of a remote control device and extracting the runtime information. The strategy is based on capturing the states of the user interface to create a model during runtime without any knowledge of the internal structure of the app. We have implemented our strategy in a tool called EvoCreeper. The evaluation results show that our strategy can automatically generate unique states and a comprehensive model that represents the real user interactions with an app using a remote control device. The models thus generated can be used to assess the quality and completeness of smart TV apps in various contexts, such as the control of other consumer electronics in smart houses.