In this talk I will present the OpenKnowledge framework and its current test applications in the case study scenarios (emergency response and bioinformatics). The framework, initially developed as a testing ground for the ideas emerging from the OpenKnowledge project, has grown into a working system that has recently seen the first release.
The framework is based on interaction models executed by peers in order to perform some coordinated task. The interaction models, written in Lightweight Coordination Calculus, describe the exchange of messages between roles taken by the involved peers, and the preconditions and effects of the messages. The models do not prescribe which peers take part in the interaction: the team is formed at run time, before the start of the interaction, trying to match the needs and capabilities of the peers.