T3: Adaptation

The ability to self-adapt will be a characteristic feature of future UC software: applications, services and system software will adapt automatically in order to offer users the optimal service. In Sub-project T3 open questions concerning the decision and control of adaptation are analyzed which have not been investigated before in the context of non-functional aspects. Thus, the adaptation constitutes a multi-dimensional optimization problem, which is based on the input from context management (T1) and knowledge discovery (T2). It must be examined whether the already used decision methods (i.e. utility functions) should be replaced by other methods (e.g. case-based reasoning or rule-based systems).

Focusing on non-functional aspects in the adaptation decision is an open question. This includes questions of human-machine interface. For example, the user must be able to influence the behavior of the system according to his current preferences. Furthermore, care must be taken to ensure that basic ergonomic rules are complied with and that the user is not confused or overtaxed by the adaptations. If a user interacts with a system in different contexts, this should not lead to an increasing learning need with regard to the user interface. From the ergonomic viewpoint it has to be investigated how a consistent interface can be guaranteed which builds on the existing practical user knowledge and conforms to the user expectations. The results of these studies are evaluated with methods of usability engineering.

Adaptation also affects the questions of confidence in such systems. How will the user be made aware that the system adapted according to his wishes? How transparent are the processes? How robustly does the adapting system work? Furthermore, adaptation generates diverse legal questions, which can affect the design of the software. The main focus is on the protection of the liability and traceability of legally relevant actions and declarations. Thus, design solutions are to be worked out for the following questions: To which degree and in which form the transparency of adaptation decisions must be guaranteed? Is it possible to increase the users’ trust and acceptance in adaptive UC systems by the development of legal rules and effective regulatory control? How about monitoring and logging the adaptation activities of a UC system?