W3C home > Mailing lists > Public > semantic-web@w3.org > November 2014

RE: Exactly one element in a RDFS class

From: Young,Jeff (OR) <jyoung@oclc.org>
Date: Mon, 17 Nov 2014 16:09:23 +0000
To: Victor Porton <porton@narod.ru>, Peter F.Patel-Schneider <pfpschneider@gmail.com>, "aldo.gangemi@istc.cnr.it" <aldo.gangemi@istc.cnr.it>
CC: Phil Archer <phil@philarcher.org>, Pavel Klinov <pavel.klinov@uni-ulm.de>, SW-forum Web <semantic-web@w3.org>
Message-ID: <257aaea7b0f54f0898ee4a2ab9316a3c@DM2PR0601MB1135.namprd06.prod.outlook.com>
If the class was treated as a blank node, it seems like that would pretty much force any individual in that graph to be a singleton of that type:

_:A0 a owl:Class .
<http://example.org/id/1> a _:A0 .

Jeff

> -----Original Message-----
> From: Victor Porton [mailto:porton@narod.ru]
> Sent: Monday, November 17, 2014 10:52 AM
> To: Peter F.Patel-Schneider; aldo.gangemi@istc.cnr.it
> Cc: Phil Archer; Pavel Klinov; SW-forum Web
> Subject: Re: Exactly one element in a RDFS class
> 
> As such I will just write:
> 
> :MyClass my:cardinality 1 .
> 
> or
> 
> :MyClass my:isSingleton true .
> 
> (I think, the first of these two signleton definitions is better, as it
> is more extensible.)
> 
> Now detecting which classes are intended to be singletons for my
> purposes is trivial.
> 
> 17.11.2014, 17:49, "Peter F. Patel-Schneider" <pfpschneider@gmail.com>:
> > Writing a SPARQL construct query to determine which classes are de
> > facto singletons is not possible, as far as I can tell.  There are
> > very many ways for an OWL class to be a de facto singleton beside
> > being equivalent to a singleton set.  For example, the class could be
> > equivalent to the intersection of two sets that have single member in
> common.
> >
> > It is also possible for non-class axioms to produce de facto
> singleton
> > OWL classes.  For example what might look to be a doubleton could be
> > turned into a singleton by a sameAs.
> >
> > In general, SPARQL is not powerful enough to analyze OWL classes.
> >
> > peter
> >
> > On 11/17/2014 07:32 AM, Aldo Gangemi wrote:
> >>  I think you need to preprocess your data with a sparql construct
> >> query to find
> >>  out what classes are de facto singletons, and to assign those
> >> classes a
> >>  punning type such as :Singleton. After that, you can use Ada.
> >>  Best
> >>  Aldo
> >>
> >>  On Monday, November 17, 2014, Peter F. Patel-Schneider
> >> <pfpschneider@gmail.com
> >>  <mailto:pfpschneider@gmail.com>> wrote:
> >>
> >>      I'm having a very hard time coming up with any overlap between
> >> this
> >>      discussion and anything that might happen in the RDF data
> shapes
> >> working
> >>      group.  The working group is about detecting explicit
> >> information in RDF
> >>      documents---this discussion is about how to create singleton
> >> classes, and
> >>      maybe how to detect such singleton classes in an RDF encoding.
> >>
> >>      That said, SPARQL is used in several of the technologies being
> >>      investigated by the working group and it is probably possible
> to
> >> write a
> >>      SPARQL query to detect a singleton class in the RDF encoding of
> >> OWL, but
> >>      this doesn't provide any true commonality.
> >>
> >>      peter
> >>
> >>      On 11/17/2014 01:50 AM, Phil Archer wrote:
> >>
> >>          This sort of debate is exactly the kind of thing that is
> >> behind the newly
> >>          formed RDF Data Shapes working group. Its charter includes
> >> pointers to
> >>          a bunch
> >>          of existing work in this area that may be useful.
> >>
> >>          See http://www.w3.org/2014/data-__shapes/
> >>          <http://www.w3.org/2014/data-shapes/>
> >>
> >>          Cheers
> >>
> >>          Phil.
> >>
> >>          On 16/11/2014 23:03, Pavel Klinov wrote:
> >>
> >>              There's no simpler encoding. Nominals is the only
> >> feature in OWL 2
> >>              which lets you say that a class has a single instance.
> >> And it has a
> >>              unique serialization in RDF.
> >>
> >>              I don't think querying for this particular syntactic
> >> construct is
> >>              complex.
> >>
> >>              However, writing RDF queries for OWL ontologies
> >> serialized in RDF (be
> >>              that SPARQL or other RDF graph matching language) is
> >> usually not a
> >>              great idea. You'll often have to deal with specifics of
> >> the RDF
> >>              serialization which is complex for many OWL constructs
> >> (see [1])
> >>
> >>              Cheers,
> >>              Pavel
> >>
> >>              [1] http://www.w3.org/TR/owl2-__mapping-to-rdf/
> >>              <http://www.w3.org/TR/owl2-mapping-to-rdf/>
> >>
> >>              On Sun, Nov 16, 2014 at 11:35 PM, Victor Porton
> >> <porton@narod.ru>
> >>              wrote:
> >>
> >>                  Your solution has the same problem as Patrick
> Logan's one.
> >>                  (See my previous
> >>                  email.) In fact your solution is the same as
> Patrick Logan's one.
> >>
> >>                  17.11.2014, 00:28, "Pavel Klinov"
> <pavel.klinov@uni-ulm.de>:
> >>
> >>                      Sorry, my previous email got sent too soon.
> >>
> >>                      Here's the link to the right place in the OWL 2
> spec:
> >>
> >>
> >> http://www.w3.org/TR/owl2-__syntax/#Enumeration_of___Individuals
> >>
> >> <http://www.w3.org/TR/owl2-syntax/#Enumeration_of_Individuals>
> >>
> >>                      Cheers,
> >>                      Pavel
> >>
> >>                      On Sun, Nov 16, 2014 at 10:52 PM, Victor Porton
> >>                      <porton@narod.ru> wrote:
> >>
> >>                             Is there any advise on how to code in
> >> RDFS or OWL
> >>                          the following statement?
> >>
> >>                             "The class X has exactly one element."
> >>
> >>                             --
> >>                             Victor Porton - http://portonvictor.org
> >>
> >>                  --
> >>                  Victor Porton - http://portonvictor.org
> 
> --
> Victor Porton - http://portonvictor.org
Received on Monday, 17 November 2014 16:09:53 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 07:42:55 UTC