Reasoning over OWL 2 is a very expensive task in general, and therefore the W3C identified tractable profiles exhibiting good computational properties. Ontological reasoning for many fragments of OWL 2 can be reduced to the evaluation of Datalog queries. This paper surveys some of these compilations, and in particular the one addressing queries over Horn-$\mathcal{SHIQ}$ knowledge bases and its implementation in DLV2 enanched by a new version of the Magic Sets algorithm.

Large-scale Ontological Reasoning via Datalog