Re: Denotation of datatype values

>On 2002-04-16 20:03, "ext Brian McBride" <bwm@hplb.hpl.hp.com> wrote:
>
>>  At 16:33 15/04/2002 -0400, Pat Hayes wrote:
>>  [...]
>>
>>>  I don't want to be a party-pooper, but I honestly feel that having an MT
>>>  and sticking to it is one way to get past this kind of half-formalized
>>>  (and rather confusing) kind of discussion. I do not know what these
>>>  'levels' are supposed to be, or how to recognize them, or how to evaluate
>>>  talk about them, etc. etc. . Why not stick to the syntax and the MT, and
>>>  just talk about that? Then everything is clear. What an application wants
>>>  to do with an RDF graph is up to it, not up to us. All we can do is to
>>>  provide application writers with a gold standard for meanings, and leave
>>>  other 'layers' to them.
>>
>>  I agree.
>
>That's a pity, because there are lots of users of RDF who can't
>read or understand the MT. So...
>
>There are many different kinds of "customers" who will read the
>RDF Datatyping specification, and we need to be sure that it is
>clear and approachable -- and ultimately *usable* -- to them all.

Agreed, but that is a matter of giving a good tutorial exposition. I 
think that can be done without mis-stating what the MT says.

>The MT, IMO, stops short of the line. It gets 95% there and fails
>to actually say how to get to the finish. It fails to capture
>the ultimate "why" of the idioms.

Patrick, you want to ALTER the MT, clearly. The trouble with your 
position, however, is that the place you want to get to, that last 
5%, is a place we have already surveyed, and we found problems with 
it. Some of our customers definitely do not want to be located there. 
They WANT to be able to be sloppy about datatype values, mix talk of 
strings with talk of integers, etc., and still they want to invoke 
lexical form checking using datatypes. They see RDF not as an 
ur-Java, or a simple data modelling language, but more like a kind of 
rather floppy semantic tagging language for attaching bits of 
half-formalized content to chunks of text. We need to be able to 
leave these sloppy Dublin guys enough flexibility to do their thing 
without having their RDF break. If we insist on the kind of tight, 
clear, sharp datatyping that is close to your (and my own) heart, 
where lexical forms *must* indicate values, then we are making RDF 
unusable by about half our user base.

>If you want the MT to be the only level addressed in the RDF
>Datatyping specification, then could we consider giving datatyped
>literal pairings a formal definition in the MT. They need no
>explicit denotation in the graph, no more so than the datatype
>mappings do. They only need definition. Eh?
>

I don't see how this will help, or even what it really means. The 
current MT is completely unambiguous, and I don't think it would be 
easy to add anything to it without breaking it somewhere. But 
chiefly, there seems to be no NEED for this. The various idioms 
provide all the precision (or lack of it) that anyone seems to need, 
as far as I  can see. If you want to talk about values, you can using 
datatype properties or dlex. If you want to insist on lexical form 
checking, you can do that. If you want to be sloppy, you can be; or 
you can state your rdfs:ranges as exactly as you want them to be if 
precision is close to your heart. Whatever makes you happy. Whats 
wrong with that?

>Patrick
>
>--
>               
>Patrick Stickler              Phone: +358 50 483 9453
>Senior Research Scientist     Fax:   +358 7180 35409
>Nokia Research Center         Email: patrick.stickler@nokia.com


-- 
---------------------------------------------------------------------
IHMC					(850)434 8903   home
40 South Alcaniz St.			(850)202 4416   office
Pensacola,  FL 32501			(850)202 4440   fax
phayes@ai.uwf.edu 
http://www.coginst.uwf.edu/~phayes

Received on Wednesday, 17 April 2002 18:02:09 UTC