SKOS/Validation
SKOS data validation and quality checking
The SKOS Reference defines a set of axioms for inference and checking integrity constraints on SKOS. However, as this paper puts it, SKOS' emphasis on minimal ontological commitment prevented the Semantic Web Deployment Group to add rules for checking SKOS data quality, which would apply to some specific domains and application only.
As a matter of fact, some patterns, which are valid according to the SKOS Core model, are discouraged in specific areas of knowledge organization (e.g., thesaurus or classification). Besides this, some of the axioms in the SKOS Reference could not be expressed in the OWL ontology language.
To palliate this, the SKOS community has come with a number of proposals for more complete data validation and quality checks. This pages lists some of these:
- Poolparty consistency checker service - rules used for consistency checking can be deduced from the quality report upon checking one vocabulary.
- qSKOS, now integrated with Poolparty. Rules have been documented in
Christian Mader, Bernhard Haslhofer, Antoine Isaac. Finding Quality Issues in SKOS Vocabularies. Proceedings of Theory and Practice of Digital Libraries (TPDL 2012), Springer-Verlag, Paphos, Cyprus, September 23-27, 2012 paper
later compared/merged with the SKOSify framework:
Osma Suominen and Christian Mader: Assessing and Improving the Quality of SKOS Vocabularies. Journal on Data Semantics, 2013. pdf
- SKOSify. Rules have been documented in:
Osma Suominen and Eero Hyvönen: Improving the Quality of SKOS Vocabularies with Skosify. Proceedings of the 18th International Conference on Knowledge Engineering and Knowledge Management (EKAW 2012), Springer-Verlag, Galway, Ireland, October, 2012. pdf
later compared/merged with the qSKOS framework
- ASKOSI performs SKOS validation using rules documented on the project's wiki
- SKOSEd uses the SKOS OWL files to do inference and validation using an OWL reasoner in Protégé
- various expressions of SKOS in OWL2 can also be used for more complete validation and inference than with the SKOS OWL 1 ontology files.