W3C home > Mailing lists > Public > public-rdf-wg@w3.org > March 2012

Re: Use cases wrt Dataset proposal (UC 1.5, UC 5.2)

From: Pat Hayes <phayes@ihmc.us>
Date: Fri, 2 Mar 2012 23:12:40 -0600
Cc: Dan Brickley <danbri@danbri.org>, RDF WG <public-rdf-wg@w3.org>
Message-Id: <05506580-79D1-485A-8CB6-1888DB1D7DC5@ihmc.us>
To: Antoine Zimmermann <antoine.zimmermann@emse.fr>

On Mar 2, 2012, at 8:09 AM, Antoine Zimmermann wrote:

> 
> 
> 
> Le 01/03/2012 20:16, Pat Hayes a écrit :
>> 
>> On Mar 1, 2012, at 2:50 AM, Antoine Zimmermann wrote:
>> 
>>> Le 01/03/2012 08:48, Dan Brickley a écrit :
>>>> 
>>>> 
>>>> On Wednesday, 29 February 2012, Antoine Zimmermann wrote:
>>>> 
>>>> [...]
>>>> 
>>>> If I say in an ontology document:
>>>> 
>>>> <>   owl:imports<http://zimmer.aprilfoolsreview.com/antoine#me>.
>>>> 
>>>> then, what I get, according to the OWL specs, is an ontology
>>>> with the content of http://zimmer.aprilfoolsreview.com/antoine
>>>> included, in spite
>>>> of<http://zimmer.aprilfoolsreview.com/antoine#me>  denoting a
>>>> foaf:Person.
>>>> 
>>>> Please take a look at the OWL 2 spec section 3.2 to check.
>>>> 
>>>> 
>>>> Why should an ontology's RDF description not include information
>>>> about real-world entities?
>>> 
>>> On the contrary, it should. It is not what I said. I say that the
>>> import mechanism does not care about what kind of entity is denoted
>>> by the IRI used with owl:imports.
>> 
>> No, it does care. If http-range-14 is correct, then that URI has to
>> denote an information resource.
>> 
>>> In the example above, the owl:imports statement implies (in the OWL
>>> 2 RDF based semantics, which is certainly the one the RDF WG is
>>> most interested in)
>>> that<http://zimmer.aprilfoolsreview.com/antoine#me>  is an
>>> ontology, but after the import is effective via a HTTP lookup, it
>>> will be inferred
>>> that<http://zimmer.aprilfoolsreview.com/antoine#me>  is a
>>> foaf:Person as well. This is consistent, as nothing says that
>>> ontologies cannot be people,
>> 
>> Actually it is impossible for an OWL ontology document (which is what
>> gets retrieved) to be a person, and there might even be true
>> ontologies out there which assert this.
> 
> Impossible?  From what do you conclude this?

Um...biology? (Do you seriously believe that a document can be a person? Even the US Supreme Court hasnt quite gone that far yet.)

> 
>> 
>> This discussion is clearly deeply involved with the whole http-range
>> issue. If we accept the TAG ruling on http-range-14, at least for
>> now, and if the http retrieval in this example returns a 200-level
>> coded response without any redirection, then we do have to assume
>> that the URI denotes an "information resource", ie a Web document,
>> more or less. And hence that it cannot be a person (or maybe better,
>> it would be false and dangerous to move ahead with any infernce that
>> was based on such a claim, as it is false and very liable to be
>> contradicted by an accepted and authoritative ontology somewhere.) So
>> I dont thikn we can be so blasé as you would like us to be about
>> situations like this.
> 
> http-range-14 has nothing to do with the semantics of OWL

No, it does precisely at this point. Http-range-14 mandates that a successful 200-coded HTTP GET result means that the URi in question is the name of an information resource. So it can't refer to something else entirely. 

> , but let me tell you that what's retrieved by the HTTP lookup in this example is the document at http://zimmer.aprilfoolsreview.com/antoine, which is *not* the URI used in the owl:imports statement.

I presume this is because  the statement has a URI reference in it, and HTTP uses the URI with the post-hash stuff removed. So yes, in this case there are two different URIrefs involved: the one that is used in the retrieval, without a hash, denotes the ontology and the one with #me added denotes you. Which is kind of odd, but does make sense. The oddity comes from using a hash URIref in an owl:imports, of course. Bad practice. 

Pat

> 
> 
>> 
>> Pat
>> 
>> 
>>> but it may seem weird to some people. But look at the following
>>> example:
>>> 
>>> 
>>> <>  owl:imports<http://xmlns.com/foaf/0.1/#stuff>  .
>>> <http://xmlns.com/foaf/0.1/#stuff>
>>> owl:sameAs<http://dbpedia.org/ontology/>  .
>>> 
>>> This will import the FOAF ontology (assuming the application is
>>> doing a HTTP lookup). However, this:
>>> 
>>> <>  owl:imports<http://dbpedia.org/ontology/>  .
>>> <http://xmlns.com/foaf/0.1/#stuff>
>>> owl:sameAs<http://xmlns.com/foaf/0.1/#stuff>  .
>>> 
>>> will import the dbpedia ontology.

Did you mean to write the second one differently? 

>>> 
>>> In spite of owl:sameAs being formally interpreted as identity,
>>> these two ontologies will eventually have very different
>>> inferences. This is because the mechanics behind owl:imports is not
>>> part of the model theory of OWL.
>>> 
>>>> 
>>>> Dan
>>>> 
>>>> 
>>>> 
>>>> 
>>>> At the very least, if we mandate this, then we need to clarify
>>>> the semantic role of "label" URIs in datasets.
>>>> 
>>>> Pat
>>>> 
>>>> ------------------------------__------------------------------
>>>> 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
>>>> <http://www.ihmc.us/users/phayes>
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>>> -- Antoine Zimmermann ISCOD / LSTI - Institut Henri Fayol École
>>>> Nationale Supérieure des Mines de Saint-Étienne 158 cours
>>>> Fauriel 42023 Saint-Étienne Cedex 2 France Tél:+33(0)4 77 42 83
>>>> 36 Fax:+33(0)4 77 42 66 66 http://zimmer.__aprilfoolsreview.com/
>>>> <http://zimmer.aprilfoolsreview.com/>
>>>> 
>>> 
>>> 
>>> -- Antoine Zimmermann ISCOD / LSTI - Institut Henri Fayol École
>>> Nationale Supérieure des Mines de Saint-Étienne 158 cours Fauriel
>>> 42023 Saint-Étienne Cedex 2 France Tél:+33(0)4 77 42 83 36
>>> Fax:+33(0)4 77 42 66 66 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
>> 
>> 
>> 
>> 
>> 
>> 
> 
> -- 
> Antoine Zimmermann
> ISCOD / LSTI - Institut Henri Fayol
> École Nationale Supérieure des Mines de Saint-Étienne
> 158 cours Fauriel
> 42023 Saint-Étienne Cedex 2
> France
> Tél:+33(0)4 77 42 83 36
> Fax:+33(0)4 77 42 66 66
> 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, 3 March 2012 05:13:18 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 16:25:47 GMT