Multi-agent based context management in AmI applications

International audienceAmbient Intelligence (AmI) is experiencing an increasing development of applications. Flexibility and genericity in the deployment and provisioning of context management solutions are key issues to be able to tackle with the large variety of domains of these context-aware applications. Existing proposals of Context Management Middleware (CMM) are still lacking such generic and flexible capabilities. To respond to these challenges, we propose CONSERT, a CMM, based on techniques and principles from the Semantic Web and Multi-Agent Systems domains. In this paper we focus on showing how the multi-agent architecture of this CMM provides the necessary flexibility to deploy different kinds of context provisioning patterns to address different AmI applications. We showcase the usage of our solution with a scenario from the domain of smart university life management