Several standards exist for expressing controlled vocabularies but with the wide adoption of the Linked Data concept, the Web-based SKOS data model has become the choice of many contributors who want to integrate their controlled vocabularies into an interconnected Web of Data.
With qSKOS it becomes possible to measure the quality of SKOS vocabularies and, most importantly, qSKOS supports thesaurus managers to achieve very high thesaurus quality. qSKOS is a set of quality issues that provide computable metrics that indicate the quality of controlled vocabularies on the Web. Given the growing number of SKOS vocabularies, such metrics can complement the intellectual quality assessment process domain experts must pass through when choosing vocabularies from the Web.
In our next major release of PoolParty Thesaurus Server (issued in June 2013) qSKOS will be fully integrated.
Thesaurus managers will have the option to execute automatic validity and quality checks like these:
- Label Conflicts (Pairs of concepts that are labeled identically) -> Assists in Finding Duplicate Concepts
- Incomplete Language Coverage (Concepts lacking documentation in a language that should be supported) -> Better performance in translation usage scenarios
- Cyclic Hierarchical Relations (Most often considered illogical and are probably mistakes) -> Better understandability
- Orphan Concepts (Concepts not semantically connected to other concepts in the thesaurus) -> Find unused concepts
- Missing Out-Links (Concepts not connected to third-party resources on the Web) -> Provide additional information without duplication
- Broken Links (References to other resources on the Web that provide no data) -> Assist in maintaining a highly informative thesaurus despite the ever-changing nature of the Web
- Full Support of the semi-formally defined integrity constraints of the SKOS reference document -> Ensures compatibility with other KOS
In addition to single checks, full reports on the thesaurus quality can be generated. Repair workflows are nicely integrated in the comfortable thesaurus editor of PoolParty. If you want to get a sneak preview of this feature, please contact our team and we will be happy to give you a demo.
qSKOS is already available on github as an open-source tool that supports checking of RDF vocabularies as input data. It is continuously maintained and improved (check out the master branch to get the stable release or the development branch for the most up-to-date code). A subset of the supported quality checks have already been published at last year’s TPDL conference.
