suggestions re ontology.owl

[VladimirAlexiev]

Suggestions re https://github.com/essepuntato/opencitations/blob/master/ontology/ontology.owl:

• is it possible to distribute/master this in turtle? easier to read
• You make statements about <https://w3id.org/oc/ontology> but that better be about <https://w3id.org/oc/ontology/> (note trailing slash) i.e. about oco: (see dct: example below).
  Same re rdfs:isDefinedBy
• vann:preferredNamespaceUri should be a string not xsd:anyURI
• use foaf: namespace
• don't define empty namespace

@prefix :      <https://w3id.org/oc/ontology/> .

You don't use it, but some triplestores will ingest it in their list of namespaces. (SKOS makes that mistake)

• don't define this namespace (unused and wrong)

@prefix skos1:    <http://www.w3.org/2008/05/skos#> .

• discrepancy:

        owl:versionIRI                 oco:1.0 ;
        owl:versionInfo                "1.1" .

• add datatype to

        dc:date                        "2016-09-23"

I made corrections like this in Turtle and sorted by subject to make reading easier. Would be glad if you use this gist

---

A general remark: I'm not sure gathering definitions of other ontologies in oco: is the best idea, some may see that as namespace hijacking. You DO change some of the definitions in minor ways, eg in original DCTERMS:

@prefix dct:    <http://purl.org/dc/terms/> .
dct:title    a rdf:Property ;
    rdfs:isDefinedBy dct: ;
    rdfs:subPropertyOf dc:title .

Your redefinition:

dcterms:title  a          owl:DatatypeProperty ;
        rdfs:isDefinedBy  <http://purl.org/dc/terms> ;

I personally find the above subProperty useless, but you get my drift.

Similarly for PROV, the original has

    rdfs:isDefinedBy <http://www.w3.org/ns/prov-o#> ;

but you write

        rdfs:isDefinedBy  <http://www.w3.org/ns/prov> ;

And you can't define which external prop connects which classes (as seen on your diagram) since redefining domains/ranges wouldn't be nice. Better use RDF Shapes to define the shape of your data.