Systematic Literature Review Protocol: Investigating Context Aware Software Testing Strategies: CAcTUS Project Report 01/2014

Ubiquitous computing shifts the focus from technology towards the user and her needs. Efforts in ubiquitous computing expand the place and mode of interaction beyond the desktop, into everyday spaces in which the main challenge is to make the systems useful in various situations that may be encountered in the real world (Lei Tang et al, 2011). A key tenant of ubiquitous computing is the concept of invisibility, where technology must blend into everyday objects (Weiser, 1991). In this context, it is essential that the different devices and services are capable of inter-operate among themselves without the interaction of the user.

With the increasing advancement in the development of ubiquitous applications, important issues related to Human-Computer Interaction (HCI) must be considered, for example: is the system adding value to the end user (Rocha, 2011), are users satisfied when using these types of applications, or whether they really met the expectations of users. These issues are even more relevant when we consider the diversity of
existing technologies (e.g., mobile phone, tablets, iPods, etc…), using different types of interaction (e.g., audio, video, text and images) between the user and the machine. In addition to this, the variety of devices poses challenges to the interoperability and the self-managing and self-organizing needs imposed by the ubiquitous environment dynamics (Rocha, 2007). We argue that interactions issues in ubiquitous systems go beyond the human computer interaction, including the interaction between different devices and systems, which we define as interaction actor-computer (between the actor and the computer), where the actor can be: a human user, another computer or device; or even another system.

In order to achieve this adaptability towards the use of these different technologies, interaction and interoperability, while minimizing the participation of the user, the ubiquitous system must be aware of its context. Software is defined as context aware when it can obtain information of the actor and the environments (computational or physical) in order to provide services or better information adapted for the user (Dey, 2001). Therefore, ubiquitous systems must be able to capture the context and use it to adapt its behavior to help the actors achieve their tasks.

Different authors have explored the problem of assuring the quality of context-aware systems (e.g., Kjeldskov et al. 2006, Fiotakis et al. 2009, O’Neil et al., 2005). In general, these studies have explored testing in the laboratory and in real environment to identify usability problems that allow the improvement of these systems in order to better serve its users. However, to better control this cycle of evaluation and
improvement, these tests must be properly designed and implemented, which leads to different research questions: what tests should be performed to ensure the best actor-computer interaction? How to consider the different possible contexts such tests? Are there methods for designing these tests that take into the context into consideration?