Right now, RDF output (both dumps and Special:EntityData) contains no reference to Wikibase ontology. This needs to be added somehow (owl:imports probably) to both, and docs/ontology.owl should have a working external URL under which it can be found.
Description
Description
Details
Details
| Subject | Repo | Branch | Lines +/- | |
|---|---|---|---|---|
| Add importing OWL ontology | mediawiki/extensions/Wikibase | master | +24 -1 |
Event Timeline
Comment Actions
wikibase:Dump rdf:type owl:Ontology . wikibase:Dump owl:imports <URI of the specific version of the Wikibase ontology used> .
Comment Actions
I personally like versioned ontology URIs (especially after the fuckup with XSD dates), but I think the consensus was to only add a version if we need one - that is, don't add a version for 1.0. See T94747: Make decision on RDF ontology prefix
Comment Actions
Setting to low- Would be nice to have, but I don't think anything is blocked on this.
Comment Actions
Change 246130 had a related patch set uploaded (by Smalyshev):
Add importing OWL ontology
Comment Actions
Copy of the question I asked on the patch:
It seems like owl:imports requires an ontology as its subject. Is it correct to claim that the dump is an ontology?
...and if we consider our data dump to be an ontology, then what isn't an ontology?
Comment Actions
Pretty much everything is, as I understand - "ontology" is the synonym for "graph" in OWL, so every OWL data set is an ontology, pretty much. But if you really feel uncomfortable with that one, I can drop the ontology part and only leave the OWL part.
Comment Actions
...and if we consider our data dump to be an ontology, then what isn't an ontology?
The word "ontology" has different meanings in different contexts. Here, we only mean the notion of "ontology" meant by the term owl:ontology as used in the W3C OWL standard. As Stas says, this is essentially nothing more than a OWL collection of OWL-compatible statements (called "OWL axioms" in the standard). No deeper meaning involved. See http://www.w3.org/TR/owl2-syntax/#Ontologies for a definition.
Comment Actions
@mkroetzsch Looking at the link, it seems to me we'd (trivially) meet these requirements. But I'm not sure about the fine details, e.g. regarding the version IRI. But if you are sure we are meeting the formal requirements, fine with me. We should then probably explicitly state that the dump is an ontology, though...
Comment Actions
Yes, that's what I meant. :-)
But I'm not sure about the fine details, e.g. regarding the version IRI. But if you are sure we are meeting the formal requirements, fine with me.
The version IRI is an optional aspect that we can include if we have a good one (I guess we might). We should give an ontology IRI and say that it is of type owl:ontology. Other bits of information about this ontology can be added, but there are not many requirements there. There is also not so much said in the standard about how to version ontologies in general, so this is something left to us.
We should then probably explicitly state that the dump is an ontology, though...
Not sure on which level you mean this. In RDF or in the user documentation? I thought that this discussion was about stating this in RDF, and this would already be "explicit". I am not sure it needs much documentation elsewhere. Mainly, we are putting this in to have a place where we can have the license (user requested feature) and other meta information (like export date and imported version of the Wikibase ontology). I agree that we should mention these features in the documentation.