RE: properties as nodes etc.

>Thanks Jonathan.
>
>>In RDF an edge corresponds not to the property or predicate URI but rather
>>to the triple itself. Many edges may share the same predicate, so the
>>predicate URI does not 'point' to or identify the edge itself.
>
>No wonder I was getting confused re. classes and instances. The primer
>doesn't help :
>
>(2.3) "In this notation, a statement is represented by a node for the
>subject, a node for the object, and a labeled arc between them for the
>predicate. ... The examples also illustrate that RDF uses URIs as predicates
>in RDF statements. "
>
>Though the MT is clearer :
>
>(0.2) "Finally, every arc in an RDF graph is labelled with a uriref. The
>same uriref may label several arcs and also be a node in the graph"
>
>So what this boils down to then (correct me if I'm wrong), to make
>assertions about a specific instance of a predicate, it's necessary to first
>reify the particular statement that contains it (which I think Seth was
>saying).

NO, that doesn't work. That makes an assertion about the piece of 
syntax, not about the instance of the property. Forget about 
reification; reification in RDF was just a mistake. The short answer 
is that there isnt any way to refer to a property instance in RDF. 
The more useful answer is that you can always introduce a subproperty 
with only one instance, and refer that instead.

>But then I don't see how information added by this means can be
>accessed without the reification, and hence what use it can be (how does an
>implementation know when to reify?).

Good question.

>
>Any clarification on this would be appreciated.
>
>>>
>>>  What I've got at the moment is something like this :
>>>
>>>     Identifiable (Temporal, Described)
>>>         |
>>>        Item
>>>     /   |   \
>>>  Vertex Edge (Adjunct)
>>>
>>>  Identifiable more or less corresponds to URI
>>
>>This is the problem, an "edge" is not a URI as opposed to vertices
>>which are
>  >named by URIs.

WHY is this perceived as a problem? What use is there in referring to edges?

Pat Hayes
-- 
---------------------------------------------------------------------
IHMC					(850)434 8903   home
40 South Alcaniz St.			(850)202 4416   office
Pensacola,  FL 32501			(850)202 4440   fax
phayes@ai.uwf.edu 
http://www.coginst.uwf.edu/~phayes

Received on Tuesday, 25 June 2002 12:10:49 UTC