Towards context-aware task recommendation

Abstract

Pervasive computing environments can provide a diverse range of computing capabilities. But they are overwhelming for us as we attempt to accomplish our daily tasks. We need to know what resources are at hand and what tasks a PCE supports. Our ongoing research aims to minimise this overload by proposing a context-aware task recommender system named TASKREC which can recommend or automatically accomplish feasible and relevant tasks for users according to user's situation and environment capabilities. This paper formalises the problem of task recommendation, presents a system architecture for TASKREC, and an early prototype implementation. Finally, we suggest open opportunities for future research on adaptation techniques for context-aware systems.