- From: <charles@w3.org>
- Date: Wed, 27 Oct 2004 06:55:41 -0400 (EDT)
- To: "Stephane Fellah" <fellah@pcigeomatics.com>
- Cc: www-rdf-interest@w3.org
Hi Stephane,
I don't see why you can't do what you want to do directly using
rdf:datatype as it is.
My understanding is that the datatype identifies a class to which the
literal belongs. In some cases you can readily describe the things in the
class (for example a thesaurus can be described as a list of things
identified by terms - the example you give) using RDF and OWL. In others,
such as when you want to relate numeric units, you need to go beyond those
two to something that can do arithmetic, although this doesn't seem
terribly complex to do.
cheers
Chaals
> Graham,
>
> Thank you for your reply. The Pan/Horrocks paper is not satisfying my
> requirements because it is focused mainly on datatypes. My proposal is
> going beyond datatypes, because it makes explicit other interpretation
> properties. You can add unit of measure or a meaning (semantic) to the
> literal (ex: some literal string may have a meaning in a taxonomy ("oak"
> in a vegetation taxonomy for example)).
>
> The canonic version of Literal as I described would be able to do this.
> My changes does not required to change the RDF Model, it just requires
> to change the representation of Literal in RDF API implementation.
Received on Wednesday, 27 October 2004 10:55:42 UTC