Model-based Data Management for Mediation Services for Intelligent Software agents

Mediation services can be generally defined as a mechanism to map interchange formats (map them to what?), thus increasing the ability for disparate systems to exchange information through common methods. However, when intelligent software agents use these meditation services, syntactical translations of formats are not sufficient. The semantic context has to be captured and interchanged as well; a common ontology is needed as the basis for the mediation service. While in the commercial world several recent publications are looking at possible automated solutions, in complex environments (is the commercial world not complex?), data engineering is necessary in order to support semantically meaningful mediation layers. Model-based data management uses a common reference model to map data models to data sources to support intelligent software agents for their internal decision processes This paper defines the phases of data engineering, shows potential conflicts and how they can be solved, and gives an example from the military application domain by showing how the Command and Control Information Exchange Data Model (C2IEDM) developed by the North Atlantic Treaty Organization (NATO) can be used as a common reference model for military applications.