W3C home > Mailing lists > Public > public-owl-dev@w3.org > July to September 2006

Re: Not out of the water yet...OWL DL properties

From: Bijan Parsia <bparsia@isr.umd.edu>
Date: Thu, 10 Aug 2006 22:55:42 +0100
Message-Id: <94BD2E64-D2B6-4430-99CC-E7A896B519B1@isr.umd.edu>
Cc: public-owl-dev@w3.org
To: Daniel Gresh <dgresh@lle.rochester.edu>

On Aug 10, 2006, at 8:11 PM, Daniel Gresh wrote:

>
> Hi,
>
> More on the image property thing. I need to give a class that  
> property,

To give a class I property (I presume you mean make a statement where  
an owl:Class is either in the subject or object position of an  
assertion) is to treat that class as an individual. That is outside  
of OWL DL (but not OWL Full). OWL 1.1 (not a standard) allows for  
"punning" that is certain semantically restricted uses of classes as  
individuals.

(OWL 1.1: <http://owl1_1.cs.manchester.ac.uk/>)

I remind folks that OWLED is coming up again. We had about 27  
submissions, but remember that the action is at the workshop itself!

> but that doesn't seem to be working out too well (it works fine for  
> Individuals).

That's inherently OWL Full.

[snip]
> And it does not work for classes.

Yep. No hope in OWL DL. You might try two files (a hack at punning),  
one that contains all the assertions using your class as a class, and  
another using it as an individual.

> It does however, work for individuals (THIS WORKS):
>
>  <j.0:InfraredFullApertureCalorimeterSystem rdf:about="http:// 
> ontology.lle.rochester.edu/IRCalSystemAppendix">
>    <j.0:ImageProperty rdf:resource="http:// 
> ontology.lle.rochester.edu/Pics/Diagnostics/Calorimetry/D-AB- 
> R-035_REVA_13.png"/>
>  </j.0:InfraredFullApertureCalorimeterSystem>
>
> Pellet says: /Multiple Types/: Resource  
> ontology.lle.rochester.edu:ImageProperty <http:// 
> ontology.lle.rochester.edu/ImageProperty> is used as an individual  
> but defined as aclass

Yep.

> *Add the following statements to make this document OWL DL*
> <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"> </ 
> rdf:RDF>
>
> That doesn't make much sense however (the add the following  
> statements part).

Ah, well, that's because you have to remove statements to make it OWL  
full. It should be clearer about that.

> So my question is: is there a way to refine the ImageProperty so it  
> can be used with classes?

Nope. Though OWL 1.1 will allow automating the "two file" solution.

> Or did I just miss some blatant error again?

Fraid so.

I believe Pellet generally tries to reason with such cases using  
punning, but that's not a portable solution.

Cheers,
Bijan.
Received on Thursday, 10 August 2006 21:55:55 GMT

This archive was generated by hypermail 2.3.1 : Wednesday, 27 March 2013 09:32:53 GMT