- From: Dave Reynolds <der@hplb.hpl.hp.com>
- Date: Fri, 01 Oct 2004 14:45:29 +0100
- To: "Miles, AJ (Alistair) " <A.J.Miles@rl.ac.uk>
- Cc: "'public-esw-thes@w3.org'" <public-esw-thes@w3.org>
[I know I said I ought to shut up but ... sorry ...] First, I think the similarity of levels is clearer if you think about things we would more normally regard as concepts, such as "person", rather than instances. So: ex:personS a skos:Concept; skos:prefLabel "person"; skos:broader ex:animalS ; skos:scopeNote "represents persons" . and ex:personR a rdfs:Class; rdfs:label "person"; rdfs:subClassOf ex:animalR ; rdfs:comment "represents persons" . seem pretty similar to me. The broader/subClassOf relations are carrying different semantics if you look in detail but the domain modeling is really much the same. I then might have: ex:resourceC ex:classifiedAs ex:PersonS; foaf:name 'Alistair Miles' . ex:resourceD rdf:type ex:personR; foaf:name 'Alistair Miles' . and indeed: ex:resourceE ex:classifiedAs ex:PersonS; rdf:type ex:personR; foaf:name 'Alistair Miles' . Now life does get harder when it comes to instances because I'm not sure what semantics are intended. If the relation was expressed as: ex:alS a skos:Concept; skos:broaderInstantive ex:personS ; skos:prefLabel 'Alistair Miles' . Then I think that is trying to express very similar intent to: ex:alR rdf:type ex:personR ; rdfs:label 'Alistair Miles' . and I don't see anything wrong at all with: ex:alS a skos:Concept; skos:broaderInstantive ex:personS; skos:prefLabel 'Alistair Miles' ; foaf:name 'Alistair Miles' ; foaf:mbox <mailto:a.j.miles@rl.ac.uk> . Whereas if we have: ex:alSG a skos:Concept; skos:broaderGeneric ex:personS skos:prefLabel 'Alistair Miles' . then presumably ex:alSG now represents "the set of resources that would be classified as Alistair Miles". In RDFS/OWL terms you might say an analogous thing using enumerated classes: ex:alRG a owl:Class; owl:OneOf { ex:alInstance } ; rdfs:label 'Alistair Miles concept' . ex:alInstance a ex:personR; foaf:mbox <mailto:a.j.miles@rl.ac.uk> . So in that case your "denotesSameAs" would link ex:alSG and ex:alRG and indeed properties like foaf:mbox would not attach directly to ex:alRG since that is the singleton class containing the instance rather than the instance itself. To me the difference is that RDFS/OWL is force you to distinguish between instantive and subclass relationships whereas skos:broader allows you to leave that open. However, that is still a matter of expressivity and detailed semantics rather some difference of metalevel. Does this make any sense? By the way, having written out this example I'd argue that if you want to build in some links between RDFS and SKOS then some official replacement for the "ex:classifiedAs" I invented above would be much higher up my priority list than "denotesSameAs". Cheers, Dave Miles, AJ (Alistair) wrote: > > > I think what I meant was ... > > If we have two resources: > > ex:resourceA a skos:Concept; > skos:prefLabel 'Alistair Miles'; > skos:scopeNote 'My mate Al.' > . > > ex:resourceB a foaf:Person; > foaf:name 'Alistair Miles'; > foaf:mbox <mailto:a.j.miles@rl.ac.uk> > . > > ... the burining questions are: > > Are these two resources fundamentally different in nature, or not? > > Are these two resources at different levels of abstraction, or not? > > My point in the last email [1] was that, it *does not* seem reasonable to > allow for example: > > ex:resourceC a skos:Concept; > skos:prefLabel 'Alistair Miles'; > skos:scopeNote 'My mate Al.'; > a foaf:Person; > foaf:name 'Alistair Miles'; > foaf:mbox <mailto:a.j.miles@rl.ac.uk> > . > > ... intuitively because properties like 'skos:scopeNote' seem appropriate to > describe an abstract concept, but not to describe a person. > > I.e. in logic speak the classes skos:Concept and foaf:Person should be > disjoint. > > What Dave said in an earlier email (I think) was that I should not think of > ex:resourceB as an actual person, but as an abstract entity, with the same > metaphysical status as ex:resourceA (i.e. they are both abstractions). > > But intuitively I do think of ex:resourceB as an actual person (should I > change that?), and that feeling is the basis for my assertion above. And > even if I do accept they are both abstractions, one *feels* to me more > abstract than the other. > > Are we getting closer or farther away? > > Al. > > [1] http://lists.w3.org/Archives/Public/public-esw-thes/2004Oct/0000.html > > --- > Alistair Miles > Research Associate > CCLRC - Rutherford Appleton Laboratory > Building R1 Room 1.60 > Fermi Avenue > Chilton > Didcot > Oxfordshire OX11 0QX > United Kingdom > Email: a.j.miles@rl.ac.uk > Tel: +44 (0)1235 445440
Received on Friday, 1 October 2004 13:45:43 UTC