Ontology-based knowledge representation for self-governing systems

Abstract

Self-governing systems need a reliable set of semantics and a formal theoretic model in order to facilitate automated reasoning. We present an ontology-based knowledge representation that will use data from information models while preserving the semantics and the taxonomy of existing systems. This will facilitate the decomposition and validation of high level goals by autonomous, self-governing components. Our solution reuses principles and standards from the Semantic Web and the OMG to precisely describe the managed entities and the shared objectives that these entities are trying to achieve by autonomously correlating their behavior. We describe how we created UML2, MOF, OCL and QVT ontologies, and we give a case study using the NGOSS Shared Information and Data model. We also set the requirements for integrating existing information models and domain ontologies into a unique knowledge base. © IFIP International Federation for Information Processing 2006.