Re: datatype properties' range

Hi Jean-Pierre,

just to make things clearer: the values of a datatype property are 
always "literals". However, in RDF, literals may have different forms:

* they can have a language tag
* they can have a datatype
* they can have none of the two above
* they can not have a language tag *and* a datatype

the last two points are sometimes considered a design default in RDF, 
and will perhaps be fixed by the current RDF WG. But for the moment, 
only "untyped" literals can have a language tag.

In OWL, the range of a datatype property specifies what kind of literal 
are expected as values of this property. If the range is a datatype 
(xsd:integer, xsd:date, xsd:string...), that will be typed literals with 
that datatype (and hence no language tag). If the range is rdfs:Literal, 
that will be untyped literals (hence possibily with a language tag).

Datatypes like xsd:int or xsd:date bring some semantics compared to 
untyped literals. However, xsd:string has no added value: it has the 
same value space as untyped literals, but does not allow language tags.

Hence my suggestion to :
- replace xsd:string by rdfs:Literal for all datatype-properties having 
xsd:string,
- keep all other range axioms untouched, as they add semantics (and 
language tags do not really make sense for integers, doubles or dates, 
anyway)

I did a quick test: datatype with range rdfs:Literal do validate OWL-RL.


   pa


On 03/09/2011 08:29 PM, Evain, Jean-Pierre wrote:
> Pierre-Antoine,
>
> I am reading a lot of material on the use of rdfs:literal combined with xml:lang. I see all sort of opinions sometimes negative.
>
> Still, I guess if we go for literal it would need to be applied more systematically and not only to strings probably using typed literals??
>
> Jean-Pierre
>
> -----Original Message-----
> From: Evain, Jean-Pierre
> Sent: mercredi, 9. mars 2011 19:31
> To: Evain, Jean-Pierre; 'Pierre-Antoine Champin'; public-media-annotation@w3.org
> Subject: RE: datatype properties' range
>
> Pierre-Antoine,
>
> If I get you right you want to type the literal with a language or...?
>
> Regards,
>
> Jean-Pierre
>
> -----Original Message-----
> From: public-media-annotation-request@w3.org [mailto:public-media-annotation-request@w3.org] On Behalf Of Evain, Jean-Pierre
> Sent: mercredi, 9. mars 2011 19:17
> To: 'Pierre-Antoine Champin'; public-media-annotation@w3.org
> Subject: RE: datatype properties' range
>
> Pierre-Antoine,
>
> About changing 'xsd:string' to rdfs:literal.
>
> In this case would it be logical to make a similar change to datatype properties with e.g. integer content?
>
> There would also likely be a need to check compliance with OWL-RL or another profile, if we care?
>
> Regards,
>
> Jean-Pierre
>
> -----Original Message-----
> From: public-media-annotation-request@w3.org [mailto:public-media-annotation-request@w3.org] On Behalf Of Pierre-Antoine Champin
> Sent: mercredi, 9. mars 2011 16:44
> To: public-media-annotation@w3.org
> Subject: datatype properties' range
>
> Hi all,
>
> in the RDF ontology, all datatype properties with textual content have
> range xsd:string.
>
> I think using rdfs:Literal instead would be beneficial, as we could
> (optionally) add a language tag to those textual values.
>
> I mention it because the API document allows to attach a language
> information to virtually any property value. This would be a nice way to
> encode this in RDF as well.
>
>     pa
>
>
> -----------------------------------------
> **************************************************
> This email and any files transmitted with it
> are confidential and intended solely for the
> use of the individual or entity to whom they
> are addressed.
> If you have received this email in error,
> please notify the system manager.
> This footnote also confirms that this email
> message has been swept by the mailgateway
> **************************************************

Received on Thursday, 10 March 2011 08:33:34 UTC