Re: SKOS revised vocabulary

Reul, Q. H. wrote:
> Hi all,
> 
> I was wondering if the latest element added to the SKOS vocabulary, such as skos:closeMatch and skos:topConceptOf, were going to be included as part of the OWL API. I seem to get a parsing problem since I have added them into my rdf file.
> 
> <skos:TopConcept rdf:about="http://www.csd.abdn.ac.uk/~qreul/thesaurus/deterioration/concepts#Deterioration_Mechanism">
>                 <skos:topConceptOf rdf:resource="http://www.csd.abdn.ac.uk/~qreul/thesaurus/deterioration">
>                 <skos:prefLabel xml:lang="en">Deterioration Mechanism</skos:prefLabel>
>                 <skos:definition>The failure mechanism (Deterioration Mechanism) causes the failure mode, which in turn causes the failure effect. Failure mechanisms provide the actual physical processes leading to a failure.</skos:definition>
>                 <skos:inScheme rdf:resource="http://www.csd.abdn.ac.uk/~qreul/thesaurus/deterioration"/>
>         </skos:TopConcept>

This example markup isn't yet wellformed XML, let alone RDF/XML.  The 
<skos:topConceptOf> element isn't closed. I suggest trying the RDF 
validator at http://www.w3.org/RDF/Validator might be of more help than 
these mailing lists. An RDF or OWL API generally can't help you if the 
files being loaded are ill-formed. First you need to make sure the file 
is well-formed XML, ie. that each tag is opened and closed 
appropriately, there are no funny characters in the wrong places. Then 
you need it to be the right kind of markup to "count" as an encoding of 
an RDF/XML graph. Only once you've got that far do things like OWL and 
SKOS make sense. And so long as you've got the RDF/XML syntax right, 
then new properties should be addable easily, and available via RDF APIs 
like Jena (RDF is quite liberal like that). I don't know the OWL API you 
mention so can't comment on those specifics.

Your example fragment would be written with ...

<skos:topConceptOf 
rdf:resource="http://www.csd.abdn.ac.uk/~qreul/thesaurus/deterioration"/>

cheers,

Dan


ps. this question would be better proposed on semantic-web@w3.org
--
http://danbri.org/

Received on Tuesday, 2 September 2008 22:53:42 UTC