W3C home > Mailing lists > Public > www-rdf-logic@w3.org > June 2002

Re: properties as nodes etc.

From: Thomas B. Passin <tpassin@comcast.net>
Date: Tue, 25 Jun 2002 17:20:21 -0400
To: www-rdf-logic@w3.org
Message-id: <002b01c21c8e$1d49ba20$961b3044@reston01.va.comcast.net>

[Danny Ayers]

> could currently be broken down something like :
>
> [sat on] --[subclass]--> [sat on for an hour]
>
> [the cat] --[sat on for an hour]--> [the mat]
>
> [sat on for an hour] --[cardinality]--> [1]
>
> but it seems less clunky to say :
>
> [sat on for an hour] --[instanceOf]--> [sat on]
>
> [the cat] --[sat on for an hour]--> [the mat]
>
> but then is [sat on] in the first figure really an arc?
>
> hey-ho ;-)
>

With Conceptual Graphs, you would say something like

[situation:
    [the cat] ->(sat on)->[the mat]
]->(duration)->[hours:1]

To do this in RDF you would have to be able to refer to a triple as a
resource.  Now since a resource can be anything identified by a URI, if we
could figure out a way to identify a triple with a URI we'd be all set.

Back to that again.

Can't there be a predicate isAnAsserted, and another resource class called
Triple?  I.e.,

{reified-statement-A,isAnAsserted,Triple}

This might be rather awkward unless a shortcut syntax were adopted, but
conceptually, why not?  Older software that did not understand the predicate
would just construct a triple, which would not be harmful, but newer
software could do something useful with it.

If we are not going to allow identification of actual triples, this approach
ought work.

Cheers,

Tom P
Received on Tuesday, 25 June 2002 17:19:25 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:52:42 GMT