W3C home > Mailing lists > Public > public-semnews@w3.org > May 2012

Re: Draft rNews 1.0 OWL Ontology: Feedback Requested [via Semantic News]

From: Ivan Herman <ivan@w3.org>
Date: Wed, 16 May 2012 13:02:20 +0200
Cc: public-semnews@w3.org
Message-Id: <2D96C534-2371-432C-9257-AA40C1CCF3FF@w3.org>
To: Stuart Myles <smyles@ap.org>
Stuart,

thanks.

I *think* there is an issue with some of the patterns you use. Consider the following:

<http://iptc.org/std/rnews/2011-10-07#encodingFormat> 
   rdfs:domain <http://iptc.org/std/rnews/2011-10-07#AudioObject>, 
               <http://iptc.org/std/rnews/2011-10-07#ImageObject> ,
               <http://iptc.org/std/rnews/2011-10-07#VideoObject> ;
 .

(I have removed the other statement on encoding format).

Let us suppose we have 

<http://www.example.org/o> iptc:encodingFormat "Something" .

An OWL reasoner will use the OWL ontology *as a license to infer* (_not_ to control the validity of the data). It will therefore infer the following three triples:

<http://www.example.org/o> rdf:type <http://iptc.org/std/rnews/2011-10-07#AudioObject> .
<http://www.example.org/o> rdf:type <http://iptc.org/std/rnews/2011-10-07#ImageObject> .
<http://www.example.org/o> rdf:type <http://iptc.org/std/rnews/2011-10-07#VideoObject> .

All these inferences are absolutely valid. I have the impression that is not what you expect...

I presume what you wanted to say is that a subject to the encodingFormat property is expected to be an audio object *or* an image object *or* a video object, but not all of these. In which case what you have to use (and I realize this is more convoluted):

<http://iptc.org/std/rnews/2011-10-07#encodingFormat>
  rdfs:domain [ a owl:Class; 
                owl:unionOf (
                   <http://iptc.org/std/rnews/2011-10-07#AudioObject>
                   <http://iptc.org/std/rnews/2011-10-07#ImageObject>
                   <http://iptc.org/std/rnews/2011-10-07#VideoObject>
                )
              ] .

In this case the OWL reasoner will infer, well, what was described in an English sentence. You may even want use:


<http://iptc.org/std/rnews/2011-10-07#encodingFormat>
  rdfs:domain [ a owl:Class; 
                owl:disjointUnionOf (
                   <http://iptc.org/std/rnews/2011-10-07#AudioObject>
                   <http://iptc.org/std/rnews/2011-10-07#ImageObject>
                   <http://iptc.org/std/rnews/2011-10-07#VideoObject>
                )
              ] .


which reinforces the fact that and audio object is not a video object (if this is also an additional feature you want to emphasize).

I hope this helps

Cheers

Ivan


On May 15, 2012, at 19:00 , Stuart Myles wrote:

> We've created an IPTC rNews 1.0 ontology file in OWL. It is available at
> http://dev.iptc.org/files/rNews/rnews_1.0_draft1.owl This is just a draft, but
> we'd like to get feedback from the experts on any improvements we can make,
> whether it is in syntax, structure, documentation or any other aspect. So, let
> us know what you think! Regards, Stuart
> 
> 
> 
> ----------
> 
> This post sent on Semantic News
> 
> 
> 
> 'Draft rNews 1.0 OWL Ontology: Feedback Requested'
> 
> http://www.w3.org/community/semnews/2012/05/15/draft-rnews-1-0-owl-ontology-feedback-requested/
> 
> 
> 
> Learn more about the Semantic News: 
> 
> http://www.w3.org/community/semnews
> 
> 
> 


----
Ivan Herman, W3C Semantic Web Activity Lead
Home: http://www.w3.org/People/Ivan/
mobile: +31-641044153
FOAF: http://www.ivan-herman.net/foaf.rdf
Received on Wednesday, 16 May 2012 10:59:14 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 19:36:31 UTC