Synthesizing probabilistic models for team-assistance in smart meetings rooms

Smart environments aim for the realisation of user assistance in an autonomously, pro-active and non-distracting manner. Therefore smart environment infrastructures needs to be able to identify users needs (intention recognition) and to be able to plan an appropriate assisting strategy (strategy synthesis). A challenge when trying to infer the users intention is the complexity of smart environments with respect to the possible tasks, number of persons and devices. We therefore examine methods to provide user assistance by synthesizing probabilistic models that are suitable to infer the state a team of users in a number of different contexts.