Re: ISSUE-30: How does SPARQL's notion of RDF dataset relate our notion of multiple graphs?

On Apr 15, 2011, at 4:18 AM, Antoine Zimmermann wrote:

> Le 14/04/2011 23:50, Pat Hayes a écrit :
>> 
>> On Apr 14, 2011, at 2:36 PM, Antoine Zimmermann wrote:
>> 
>>> Our decision today was about SPARQL datasets and what the URI in
>>> the <n,G>  pairs is referring to. We said it does not necessarily
>>> identify the graph in the sense of what the RDF semantics says (the
>>> interpretation of n does not need to be the graph). This is enough
>>> to define a notion of interpretation and model of a dataset, as
>>> explained in section "Interpreting datasets" of
>>> http://www.w3.org/2011/rdf-wg/wiki/TF-Graphs/RDF-Datasets-Proposal.
>> 
>>> 
>> Oh, wait. We did NOT agree to accept an alien model theory for
>> datasets. If you wish to propose this, please make it an action item
>> for the WG. I will argue strongly, and vote, against it. But in any
>> case, the proposed model theory for datasets, in the proposal, uses
>> the same language or reference in an interpretation, so does not
>> resolve the issue I was pointing out.
> 
> The proposed semantics for datasets does not forbid defining a mechanism to identify graphs. What Pierre-Antoine describes in his email would work, I believe.
> 
> What is written in "Interpreting datasets" is probably the least constrained form of semantics you can define for a dataset, based on RDF interpretations. It fulfils the need for having diverging opinions expressed in different graphs (or even contradicting graphs) and if some people want more constraining semantics, it's not a problem to extend it.

It is a problem; but in any case, we have taken the decision that URI<>graph 'tagging' is *not* related to any model-theoretic notions of reference, so appealing to an interpretation-based semantics is pointless to address the issue. With the current decision, the use of a URI to 'identify' a graph in a SPARQL dataset is *necessarily* unrelated to *any* use of URIs to name graphs defined in any model-theoretic semantics. 

> If we don't define the semantics, I'm afraid there's going to be endless discussions or misunderstanding about what the stuff in the dataset entails.

I entirely agree. I wish we had defined the semantics of datasets first, in fact. 
> 
>>> 
>>> Now, the use cases clearly show that we need a way to identify a
>>> graph (or rather a g-box) with a URI. My understanding was that
>>> this is independent from today's decision, and I hope it is.
>> 
>> Quite. And if it is, then this use of a URI to identify a graph
>> (g-box, whatever) will be independent of the SPARQL use of URIs to
>> identify graphs. And hence, my point about the use of URIs in RDF
>> triples will still stand.
> 
> I do not see what is the problem of having URIs that identify graphs (in the RDF interpretation sense) together with additional labels for graphs which just happen to be URIs too.

That is not the point. Of course URIs can be used for many purposes. 

> 
> :antoine { :g  a  :Graph . :antoine a :Person }
> 
> works for me. If I want I can (for my personal purpose, not for the WG) extend the RDF interpretation with an additional "graph-map" which maps URIs to graphs such that a URI can identify a thing (a person, a cow, an idea) as well as a graph.

I don't believe you can, in fact. That is, you cannot write a coherent model theory using such 'mappings'. 

> In OWL 2 DL, you have 3 interpretation functions that maps URIs to different things. An individual-interpration, a property-interpretation and a class-interpretation. These three interpretations can map the same URI to 3 distinct things, such that a URI "names" multiple things having a different nature.

Yes, I know. And in the Common Logic semantics, one 'thing' can have multiple extensions associated with it: the same name can denote an individual, function, relation, class and proposition. But this (either punning or what has been called 'pollarding') works only under special circumstances, to wit, that the appropriate meaning of the URI or name can be determined by the *syntactic* context in which the name occurs.  Without this, the language is ambiguous and it is impossible to give a coherent semantic construction. (Believe me, I have tried. We tried to extend the ISO Common Logic to allow such name-ambiguity, and the result was horrendously complicated and eventually was abandoned.) It is not a license to allow arbitrary multiple-uses of names. 

> 
> Names map to things *in a context*.

If you mean a syntactic context, then yes, but RDF has very limited syntax to define such 'contexts' (unlike OWL or CL), and it is already using them in its present semantics. If you mean some more exotic notion of context, then I disagree. Names do not map to things in contexts, at least in any conventional logic. They map to one thing per interpretation. 

> A URI names a thing in RDF by way of its RDF-interpretation but a URI can name a graph in a dataset by way of what is called "named graphs" in the SPARQL spec.

I dont think this claim makes sense. My point is that if the URI is used in an RDF triple, then what it names must be specified using the RDF semantics. Which can be done, quite easily: but we have decided, however it is done, that shall **not** be related to the use of any URI to be a 'tag' for a graph in a SPARQL dataset. And it is that decision to not relate them which is the issue here.

> I could also name my dog Tim Berners-Lee, or even name it using TimBL's URI and register this identification somehow and that would not be a problem for RDF (although, I agree, it would be silly).

It would be a problem if RDF treated that identification as a logical naming, ie if it were required to hold in all satisfying interpretations (as the named graph syntax requires.) 

> 
> 
> AZ.
> 
>> So all is not good, unless you are in agreement that SPARQL cannot
>> use the same identifier in an RDF object position and also to
>> identify a graph in a dataset.
>> 
>> Pat
>> 
>>> 
>>> So all is good so far.
>>> 
>>> 
>>> Le 14/04/2011 19:09, Pat Hayes a écrit :
>>>> Well, the use of a URI inside an RDF triple assumes that the URI
>>>> is being used as a name, to refer to something. Using a URI which
>>>> is the name of a named graph, for example, would refer to the
>>>> graph. But in this decision we *explicitly* say that this is
>>>> *not* how the SPARQL association of URIs to graphs works: that
>>>> the 'associated' graph which is 'tagged' (if I have that right)
>>>> by a URI might well not be the entity referred to by the URI. The
>>>> example was given in which the URI is the name of a person, ie
>>>> refers to a person, and still can be used to 'tag' a graph for
>>>> SPARQL purposes. If such a URI is used as the object of an RDF
>>>> triple, it will refer to the person, not to the SPARQL-tagged
>>>> graph. As there is no way to know whether the graph that is
>>>> SPARQL-tagged by a URI is, or is not, the referent of the URI,
>>>> any use of that URI as a name inside an RDF triple must be
>>>> basically unrelated to its use as a SPARQL graph tag; or at any
>>>> rate, that is the only safe assumption to make.
>>>> 
>>>> In a nutshell, RDF uses URIs as referring names. Apparently,
>>>> SPARQL does not, when it comes to identifying graphs. So the uses
>>>> of URIs in RDF triples and in SPARQL tags are dissociated from
>>>> one another, and need have no relationship. So, no relationship
>>>> can be relied upon. The 'naming' of graphs in SPARQL is a wholly
>>>> SPARQL-local business, unrelated to RDF semantics and therefore
>>>> to any RDF content.
>>>> 
>>>> I assumed this was obvious at the time we were discussing this,
>>>> by the way. But I confess I had not at that time read the Wiki
>>>> proposal fully, and not seen the 'imports' examples.
>>>> 
>>>> Pat
>>>> 
>>>> On Apr 14, 2011, at 11:55 AM, Ivan Herman wrote:
>>>> 
>>>>> Pat,
>>>>> 
>>>>> sorry, but you will have to explain (me) what the problem is.
>>>>> 
>>>>> Ivan
>>>>> 
>>>>> ---- Ivan Herman web: http://www.ivan-herman.net mobile: +31
>>>>> 64 1044 153
>>>>> 
>>>>> On 14 Apr 2011, at 18:43, Pat Hayes<phayes@ihmc.us>   wrote:
>>>>> 
>>>>>> I note in passing that the Proposed WG Decision dated 14
>>>>>> April has the consequence that the IRi associated with a
>>>>>> graph in SPARQL cannot be used inside an RDF triple to
>>>>>> reliably refer to the graph. This means in particular that
>>>>>> uses such as those contemplated in
>>>>>> http://www.w3.org/2011/rdf-wg/wiki/TF-Graphs/RDF-Datasets-Proposal,
>>>>>> 
>>>>>> 
> which use the SPARQL name as the object in an 'imports' triple,
>>>>>> are ruled out by this decision.
>>>>>> 
>>>>>> Pat
>>>>>> 
>>>>>> On Apr 13, 2011, at 4:29 AM, RDF Working Group Issue Tracker
>>>>>> wrote:
>>>>>> 
>>>>>>> 
>>>>>>> ISSUE-30: How does SPARQL's notion of RDF dataset relate
>>>>>>> our notion of multiple graphs?
>>>>>>> 
>>>>>>> http://www.w3.org/2011/rdf-wg/track/issues/30
>>>>>>> 
>>>>>>> Raised by: On product:
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>> 
>>>>>> ------------------------------------------------------------
>>>>>> IHMC (850)434 8903 or (650)494 3973 40 South Alcaniz St.
>>>>>> (850)202 4416   office Pensacola (850)202 4440   fax FL
>>>>>> 32502 (850)291 0667   mobile phayesAT-SIGNihmc.us
>>>>>> http://www.ihmc.us/users/phayes
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>> 
>>>> 
>>>> ------------------------------------------------------------
>>>> IHMC (850)434 8903 or (650)494 3973 40 South Alcaniz St. (850)202
>>>> 4416   office Pensacola                            (850)202 4440
>>>> fax FL 32502                              (850)291 0667 mobile
>>>> phayesAT-SIGNihmc.us       http://www.ihmc.us/users/phayes
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>> 
>>> 
>> 
>> ------------------------------------------------------------ IHMC
>> (850)434 8903 or (650)494 3973 40 South Alcaniz St.
>> (850)202 4416   office Pensacola                            (850)202
>> 4440   fax FL 32502                              (850)291 0667
>> mobile phayesAT-SIGNihmc.us       http://www.ihmc.us/users/phayes
>> 
>> 
>> 
>> 
>> 
> 
> 
> -- 
> Antoine Zimmermann
> Researcher at:
> Laboratoire d'InfoRmatique en Image et Systèmes d'information
> Database Group
> 7 Avenue Jean Capelle
> 69621 Villeurbanne Cedex
> France
> Tel: +33(0)4 72 43 61 74 - Fax: +33(0)4 72 43 87 13
> Lecturer at:
> Institut National des Sciences Appliquées de Lyon
> 20 Avenue Albert Einstein
> 69621 Villeurbanne Cedex
> France
> antoine.zimmermann@insa-lyon.fr
> http://zimmer.aprilfoolsreview.com/
> 

------------------------------------------------------------
IHMC                                     (850)434 8903 or (650)494 3973   
40 South Alcaniz St.           (850)202 4416   office
Pensacola                            (850)202 4440   fax
FL 32502                              (850)291 0667   mobile
phayesAT-SIGNihmc.us       http://www.ihmc.us/users/phayes

Received on Saturday, 16 April 2011 04:07:45 UTC