- From: Jeremy Carroll <jjc@hpl.hp.com>
- Date: Thu, 19 Dec 2002 14:49:51 +0100
- To: www-webont-wg@w3.org
I should take back my original comment about owl:Ontology. (it's too late isn't it?). I had misunderstood the # in the rules. # in N-triples is the comment character, so I had read those productions has having no real effect!! (Dan's suggestion of <> would be better). The <> rdf:type owl:Ontology . triples could be generated by any of the following: 1: using Peter's new rules. 2: using the original annotation rule Annotation( <URI reference> <URI reference> ) => # <URI reference> <URI reference> . with appropriate URIrefs (which would have to be permitted explicitly; presumably the permitted annotations in the owl namespace will need to be explicitly listed in the abstract syntax). 3: using a modified annotation rule: Annotation( <URI reference> <URI reference> ) => # <URI reference> <URI reference> . # rdf:type owl:Ontology . i.e. that owl:Ontology is required iff there is at least one other Annotation. Impact of [1] is that <owl:Ontology rdf:about=""/> is obligatory - this is a change from D+O. Impact of [2] is that <owl:Ontology rdf:about=""/> is optional. Impact of [3] is that the document info section as a whole is optional, but must use owl:Ontology and must not be empty. The modified rules look much better. I note there are still a few bugs: e.g. no rule for empty sequences (main node is rdf:nil). I think the Class rule should be split into three cases: - partial with no descriptions - with one description - with two or more descriptions This could avoid the ugly spurious rdf:Description owl:intersectionOf elements in say: http:../../www-archive/2002Dec/att-0071/00-t#proposedRDFXMLFunction-cardinal ity D.1.6 owl:cardinality (quoted at end of message) I think the use of both annotation and Annotation as two distinct syntactic constructs is poor naming - how about "OntologyAnotation" and "annotation"? The abstract syntax should require at least one description in complete Class constructs. The cardinality rules currently require the use of xsd:decimal in owl Lite and OWL DL. It might be better to: a) use xsd:nonNegativeInteger b) explicitly permit any type derived by restriction from xsd:decimal as long as the value is in the value space of xsd:nonNegativeInteger While this *is* non-deterministic, I note that the key use of the abstract syntax in the spec is to define OWL Lite and OWL DL which are syntactic subsets of OWL Full. Jeremy <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#" xmlns:owl="http://www.w3.org/2002/07/owl#" xml:base="http://www.w3.org/2002/03owlt/cardinality/premises001" > <owl:Class rdf:ID="c"> <rdfs:subClassOf rdf:resource="http://www.w3.org/2002/07/owl#Thing" /> <rdfs:subClassOf> <rdf:Description> <owl:intersectionOf rdf:parseType="Collection"> <owl:Restriction> <owl:onProperty rdf:resource="#p"/> <owl:cardinality rdf:datatype="http://www.w3.org/2001/XMLSchema#nonNegativeInteger" >1</owl:cardinality> </owl:Restriction> </owl:intersectionOf> </rdf:Description> </rdfs:subClassOf> </owl:Class> <owl:ObjectProperty rdf:ID="p"> <rdfs:range rdf:resource="http://www.w3.org/2002/07/owl#Thing"/> <rdfs:domain rdf:resource="http://www.w3.org/2002/07/owl#Thing"/> </owl:ObjectProperty> </rdf:RDF>
Received on Thursday, 19 December 2002 08:50:12 UTC