W3C home > Mailing lists > Public > semantic-web@w3.org > September 2006

Re: Inheriting property restrictions

From: ben syverson <w3@likn.org>
Date: Wed, 6 Sep 2006 20:45:58 -0500
Message-Id: <C83128DB-9581-4530-BD1E-14BC27F1777B@likn.org>
To: semantic-web@w3.org

Thanks Paul!

I guess it's as I feared -- multiple property restrictions result in  
the intersection of the restrictions, rather than the union.

No big deal -- I just have to make this extremely clear to users so  
they don't create an impossible (or even improbable) intersection of  
restrictions, thereby preventing them from actually applying useful  
values to the property.

Maybe this should be clearer in the OWL specs? It doesn't really talk  
about multiple property restrictions and how they interact...

- ben syverson
likn.org



On Sep 6, 2006, at 5:09 PM, Paul Gearon wrote:

>
> Hi Ben,
>
> The quick answer is that rdfs:subClassOf is a transitive relationship.
>
> To see how that applies to your case, it means that WebCar is a
> subclass of Car, the Colors restriction, and the HTMLColors
> description.  So both the restrictions must be applied.
>
> As to the meaning of this... it depends on whether Colors and
> HTMLColors are mutually exclusive sets.  If not, then there's no
> issue.  If they are (as I infer here), then it is not possible to
> create a statement which uses the PaintColor property on a WebCar
> (incidentally, the convention is that properties start with a lower
> case letter).  It's perfectly legal to be unable to create a statement
> like this, unless you've declared a minimum cardinality for PaintColor
> that is greater than 0.
>
> On your second question, it is fine to add more restrictions to the
> use of a predicate in a subclass, so it should not generate a warning
> in the general case.
>
> OTOH, if it can be determined that the set of values is an empty set
> (for instance, several owl:allValuesFrom restrictions with an empty
> intersection) then it might be useful to generate a warning.  However,
> determining this will be extremely hard or impossible except in simple
> cases, so I wouldn't think it is something you can do most of the
> time.
>
> Regards,
> Paul
Received on Thursday, 7 September 2006 01:46:11 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 07:41:53 UTC