Agreement technologies aim at developing techniques that enable software components to reach agreements on the mutual performance of services, hence supporting the development of large-scale, open distributed systems. Hence, agreement technologies propose a new paradigm for next generation distributed systems. The new paradigm is structured around the concept of agreement between computational agents. These agreements must be consistent with the normative context where they are established and permit, once accepted, that agents call for mutual services and honour them. An entity, by the fact of being autonomous, may choose whether to fulfil an agreement or not, and it should fulfil them when there is an obligation to do so derived from the standing agreements. Autonomy, interaction, mobility and openness are the characteristics that the paradigm covers from a theoretical and practical perspective. We have identified negotiation, semantic alignment and trust as the key enabling agreement technologies.
- Negotiation techniques allow agents to reach agreements on the terms of their interactions.
- Semantic alignment techniques allow agents using different ontologies to understand one another.
- Trust is the technology that complements traditional security mechanisms by relying on social mechanisms that interpret the behaviour of agents.
Semantic ---in addition to syntactic--- heterogeneity is an inherent problem for achieving interoperability in distributed systems if system components have been engineered separately and autonomously ---as in federated databases, peer-to-peer networks, choreographed web services, or multiagent systems. To guarantee an acceptable degree of interoperability in any of these systems, some sort of agreement on the semantics of the representation of the application domain is needed.
We have been investigating how to tackle semantic heterogeneity with the objective of reaching semantic agreements by means of semantic alignment, understood in a very broad sense. Hence, our research focusses on:
- mathematical foundations of semantic interoperability and integration
- ontology matching and mapping
- meaning coordination and negotiation
- interaction-situated semantic alignment
- alignment of subjective evaluations of trust