We study the problem of approximating Description Logic (DL) ontologies specified in a source language LSLS in terms of a less expressive target language LTLT. This problem is getting very relevant in practice: e.g., approximation is often needed in ontology-based data access systems, which are able to deal with ontology languages of a limited expressiveness. We first provide a general, parametric, and semantically well-founded definition of maximal sound approximation of a DL ontology. Then, we present an algorithm that is able to effectively compute two different notions of maximal sound approximation according to the above parametric semantics when the source ontology language is OWL 2 and the target ontology language is OWL 2 QL. Finally, we experiment the above algorithm by computing the two OWL 2 QL approximations of a large set of existing OWL 2 ontologies. The experimental results allow us both to evaluate the effectiveness of the proposed notions of approximation and to compare the two different notions of approximation in real cases.
|Title of host publication||The Semantic Web - ISWC 2014|
|Subtitle of host publication||13th International Semantic Web Conference, Riva del Garda, Italy, October 19-23, 2014. Proceedings, Part II|
|Publisher||Springer International Publishing|
|Number of pages||16|
|Publication status||Published - 2014|
|Name||Lecture Notes in Computer Science (LNCS)|
|Publisher||Springer International Publishing Switzerland|