A classification and modeling of the quality of contextual information in smart spaces

Reliable contextual information should be generated to provide pervasive services to the occupant in smart spaces. This is difficult for several reasons. First, the number of ways to describe an event or an object is unlimited and there is no standard regarding granularity of context information in context classification schemes. Second, the quality of a given piece of contextual information is not guaranteed by uncertainty. In this paper, we propose a pragmatic context classification and a generalized context modeling scheme based on sensor fusion techniques. To make a pragmatic context classification, we introduce two approaches, “occupant-centered pragmatic approach” and “relation-dependency” approach. To improve the quality of given contextual information by reducing uncertainty, we introduce “state-space based sensor fusion modeling” as a generalized context modeling. Finally, we show an example within the applied scenario as an evidential network.