Contexts and context-awareness revisited from an intelligent environments perspective

Context is a useful concept somehow unconsciously used by humans in daily life problem-solving. Recently several subareas of computer science have been increasingly trying to rely on this concept to design systems with practical use in certain predefined circumstances. The perception is that imbuing in the system certain context-awareness qualities can support intelligent decision-making in specific practical situations. Despite a significant number of implemented systems which aim at providing context-awareness there is a lack of commonly accepted and used methodologies and tools. At the root of this, is the lack of agreement on a set of good principles or standards which can act as a guide to the scientific community and the developers interested in this class of systems. There have been some extensive surveys on the use of context, still there is no theoretical corpus emerging which we can use to discuss the essential concepts making up the fabric of contexts and its use by system developers. Here we attempted such enterprise at a level which is more formal than popular surveys, in a way that is not implementation dependent and in a way that highlights key concepts of relevance to developers. We reassessed first the basic concepts identifying the need to more prominently consider system beneficiaries’ satisfaction. We then transfer explicitly these values to a more formal outline of the basic components and the operations which emerge as relevant. We identify and highlight the tasks of context activation, comparison, influence, construction, and interaction. We hint at how these may work in practice and explained these through examples. We show how the theory is flexible enough by generalizing it to multiusers so that optimization of global preferences and expectations is used to drive system development and system behaviour