An autonomic system to characterize the physical and organizational context of public spaces

Traditionally, context-aware applications are developed using one of the following approaches: each application communicates directly with sources of context information; applications are developed with the aid of reusable libraries for processing context information or applications have their own context models. These approaches not only require effort on the part of those who develop them, but also create dependency between the developed context-aware applications and the systems developed for obtaining and providing the context information, thus hindering their reuse in other applications. In this paper we present a system that autonomously characterizes the physical, social and organizational context of a public space and delivers this information to context-aware applications. Experiments show that it provides relevant and useful context information that can be used by context-aware applications to improve their services.