Modeling ontologies as executable domain specific languages

Ontologies, one of the cornerstone technologies of the Semantic Web are a very expressive and powerful means for domain modeling. To become really usable and ubiquitous, the implementations of ontology languages yet have to deliver satisfactory performance and before anything else have to be developer-friendly. We present an example of using Magic Potion, a description logics-based domain specific meta-language for building executable domain models. We are using LOCO ontology framework developed to support personalised learning contexts in learning applications to illustrate how ontologies in the form of executable domain-specific languages can blend the Semantic Web ideas into classical software engineering to improve expressiveness of domain models.