Owen Conlan, Ruaidhrí Power, Steffen Higel, Declan O'Sullivan, Keara Barrett

Next Generation Context Aware Adaptive Services

Knowledge and Data Engineering Group
Department of Computer Science
Trinity College Dublin and
Telecommunications Software and Systems Group (TSSG)
Waterford Institute of Technology (WIT)

Situational information can enrich the interactions between a user and the services they wish to utilize. Such information encompasses details about the user, the physical environment and the computing resources. There are at least three key aspects in addressing this issue. Firstly, it is important to accurately capture or infer the requirements of the users in a timely fashion. Without precise information on what the users are hoping to achieve it is difficult to identify suitable services or sub-services that may fulfill (in part or fully) their information needs. This information may be obtained by combining information inferred about the users with that entered through direct user control mechanisms, thus empowering the users. Secondly, the nature of the available services determines the modes in which they may be adapted to the users’ needs. Rigid, inflexible services may be difficult to tune to the information requirements of the users. Adaptive services, on the other hand, are well suited to dynamically modifying their behavior, within defined constraints. The third issue to be addressed is the on-the-fly combination of services to meet the users' requirements. A single service attempting to fulfill all of the users’ requirements may suffer from a bloating problem. Adaptive services, coupled with information about the users, are central to the ability to assemble ad-hoc conglomerate services from sub-services. The service delivering information to the user must, however, be able to utilize this information effectively to contextualize and adapt its output to the user. Adaptive services should be able to react to the user's surroundings as well. By combining adaptive services the potential for utilizing context information is increased. Finally, an adaptive service should be conscious of the computing context in which it is operating. This paper argues that current modeling (both of users and services) techniques, adaptive axes and personalization techniques used in current personalized information services, such as Adaptive Hypermedia Systems, may supply the basis for next generation adaptive collaborative services.

Keywords: Adaptivity, Modeling, Context, Services.