Re: PROV-ISSUE-142 (Tlebo): Can roles only be Literals? [Data Model]

On Nov 10, 2011, at 10:02 AM, Graham Klyne wrote:

> On 07/11/2011 22:19, Luc Moreau wrote:
> > Here is a first draft of the literal section.
> >
> > http://dvcs.w3.org/hg/prov/raw-file/default/model/ProvenanceModel.html#record-literal 
> 
> I have problems with this.
> 
> (1) The first paragraph is confusing.  Es[pecially the second sentence - what is this trying to say?
> 
> (2) "a PROV-DM Literal may be a URI-typed string (with datatype xsd:anyURI), or URI-denoted resource (with datatype rdf:Resource); in either case; such URI has no specific interpretation in the context of PROV-DM."
> 
> Firstly, the structure here is so close to the structure of RDF literals, that to then not adopt the standard RDF semantics for them is, IMO, a recipe for much confusion.
> 
> Secondly, the idea of using rdf:Resource as a datatype URI seems to fly in the face of RDF.  If you are going to explicitly use RDF URIs, I think it's crucial to adopt RDF semantics for them.
> 
> ...
> 
> Personally, I think DM would do better to back off from specifying URIs and RDF-style literals, and just talk about names and literals, specifying a minimum that you need to express the abstract model.


+10

We seem safe with "A PROV-DM Literal represents a value whose interpretation is outside the scope of PROV-DM."
and
"in either case; such URI has no specific interpretation in the context of PROV-DM."

-Tim


> 
> The Ontology document can then carry the task of mapping the abstract model to RDF, which IMO would lead to a much cleaner separation of concerns.
> 
> #g
> --
> 
> 
> On 07/11/2011 22:19, Luc Moreau wrote:
>> Hi Tim, Stephan, Jim,
>> 
>> Here is a first draft of the literal section.
>> 
>> http://dvcs.w3.org/hg/prov/raw-file/default/model/ProvenanceModel.html#record-literal
>> 
>> 
>> It would be good to have your feedback.
>> If you find it's ok, than the literals examples in the document need to be checked.
>> 
>> Cheers,
>> Luc
>> 
>> On 07/11/11 18:15, Jim McCusker wrote:
>>> On Mon, Nov 7, 2011 at 12:42 PM, Paolo Missier<Paolo.Missier@ncl.ac.uk> wrote:
>>>> DM says:
>>>> 
>>>> 5.5.5 Literal
>>>> 
>>>> Literals represent data values such as particular string or integers.
>>>> 
>>>> My understanding is it's always been used in the standard grammar production
>>>> meaning (eg: http://savage.net.au/SQL/sql-2003-2.bnf.html#literal). Not so?
>>> I think a clearer definition would be:
>>> 
>>> A Provenance Literal is a "leaf" value. It does not explicitly have
>>> any outgoing relations (in SW-ish: Is not a subject of any statement)
>>> in the PROV data model. Any outgoing relations from a Provenance
>>> Literal is out of scope for the PROV DM.
>>> 
>>> Jim
>> 
> 
> 

Received on Tuesday, 22 November 2011 17:50:09 UTC