RE: rdfs:class and rdfs:resource

On Mon, 2003-05-12 at 10:17, LYNN,JAMES (HP-USA,ex1) wrote:
> This brings up an important issue. I don't believe there is any mechanism in
> RDF/RDFS for expressing notions of "time-varying properties" That's not too
> say that it cannot be expressed with the current specs, rather I should say
> it is not a specified property. At the risk of adding one more thing to the
> list of "things which need to be ruled upon" (like "who owns URIs" and
> "social meaning"), does this need to be part of the next spec? I'm sure
> everyone has their own favorite examples, but just as a starting point (sans
> RDF):
> 	"John Doe is in Paris."
> 
> Is this true? To make it unambiguous, do I need to specify a time range? If
> I don't know the time range, can I use a relative term such as "during the
> time I was in Paris" ?

It might help to think about the statement as a noun.  Something like:

John Doe has Travel Itinerary.
Travel Itinerary has Layover.
Layover hasCity Paris.
Travel Itinerary startsOn "June 5".
Travel Itinerary endsOn "June 8".
Layover startsOn "June 5, midnight".
Layover endsOn "June 6, 3 am".

So, instead of saying "John Doe is running." you can say "John Doe had
Run."  By taking the action (or the "is" in your above example) and
making it a Thing (noun), you can assign all sorts of properties to it. 
Such as startsOn or endsOn, to take care of the time issue.

I can elaborate if that helps,
Seth

Received on Monday, 12 May 2003 11:33:37 UTC