> Max V=F6lkel wrote:
> >>> on a page "Joe":
> >>> [[knows :: [[who :: Frank, kind :=3D "friend", since :=3D 1992]]
> >>> :: [[who :: Bob, kind :=3D "acquaintance", since :=3D 1981]]
> >> What are the intended semantics of the above example?
> >>in n-triples notation:
> >>:Joe :knows _:1.
> >>_:1 :who :Frank
> >>_:1 :kind "friend"
> >>_:1 :since "1992"
> > Correct.
> > The presentation of such semantic info in the fact-box would change
> > it does not show things like a href to _:1, but instead shows a
> > "nested" fact box about _:1
> Hmmm, this is pretty weird modeling.
> Whats an intuitive way of modeling "Joe knows Frank, they are friends
> since 1992" in OWL?
I guess, by giving _:1 a more meaningful name:
(Joe, Relation:knows, Frank)
(Joe, _:is_since_1992_friend_of, Frank)
(Frank, _:is_since_1992_friend_of, Joe) // you might want to use
(_1, rdf:type, owl:SymmetricProperty) instead
(_:is_since_1992_friend_of, Attribute:since, 1992)
(Relation:knows, rdf:type, owl:ObjectProperty)=09
(Relation:is_friend_of, rdfs:subPropertyOf, Relation:knows)
(Attribute:since, rdf:type, Type:DateTime)
(Type:DateTime, rdf:type, rdfs:Literal)
I'm not sure if its 100% valid, because I've intuitively* modeled it ;-) =
* =3D I think that doesn't count much, because I know better how to =
Ontologies than the common.