Search Mailing List Archives
[protege-owl] Define more than Subject-Predicate-Object
tredmond at stanford.edu
Sun Apr 29 22:47:34 PDT 2007
On Apr 27, 2007, at 1:30 PM, Steve Jonathan Tirtha wrote:
> After studying OWL, I see, that OWL can only represent a statement/
> knowledge in S-P-O(Subject-Predicate-Object).
> Can I also define S-P-O-Context. I don't know the right term, so I
> call it "Context".
> Example: S-P-O
> - Person signs Contract
> - Michael signs LicenseAgreementNr.12
> - Person signs Contract on [Time Context]
> - Michael signs LicenseAgreementNr. 12 on August
> 12th 2006
> [Time Context] defines when a person signs a contract.
I think that the issue you are getting at is the problem described in
http://www.w3.org/TR/swbp-n-aryRelations/. This is a known problem
and that web page gives a couple of approaches for solving the problem.
> I think, OWL can't represent more than S-P-O.
This is a little subtle. The statement you made is quite reasonable
for the RDF language. So one could argue that since OWL has a
semantics based on the RDF semantics (http://www.w3.org/TR/owl-
semantics/rdfs.html) the statement that you made also applies to
OWL. This is a defensible position.
However I would argue that this is not an entirely natural
viewpoint. Say we look at a statement like
MyFavoritePizza = Pizza and (hasTopping some SausageTopping) and
(hasBase all ThinAndCrispyBase).
It is possible to express the semantics of such a statement in terms
of triples. But I would argue that the more natural semantics is the
direct model theoretic semantics (http://www.w3.org/TR/owl-semantics/
direct.html). There is no need to turn this statement into a
collection of triples in order to understand its meaning.
More information about the protege-owl