Ontologies, Meta-models, and the Model-Driven Paradigm

So far, ontologies in the Semantic Web and models in model-driven engineering have been developed mainly in isolation. It seems that due to a lack of communication between communities, modelling concepts have been designed similarly in both paradigms without ensuring their orthogonality. On the long run, this will replicate efforts and cannot be afforded by either community. Hence, this chapter discusses the role of ontologies, models, and meta-models in the model-driven engineering (MDE). To show how ontologies can be employed in MDE, in particular, in its variant model-driven architecture (MDA), the chapter presents a meta-modelling hierarchy that is aware of ontologies—that is, an ontology-aware mega-model of MDE. Based on the insight of [38] that the main difference of models and ontologies lies in their descriptiveness resp. prescriptiveness, the role of ontologies in this meta-pyramid is to describe the existing world, the environment, and the domain of the system (analysis), while the role of system models is to specify and control the system under study itself on various levels of abstraction (design and implementation). Consequently, in this scheme, MDE starts from ontologies, refines, and augments them towards system models, respecting their relationships to prescriptive models on all metalevels.