Re: Are literals owl:Things?

Jeremy,


Le 14/10/2010 17:41, Jeremy Carroll a écrit :
> Yes
>
>
> The way to answer this is to assess the truth of
>
> rdfs:Literal rdfs:subClassOf owl:Thing .


Well, this is one way of answering the question. I would say, this is 
the "proof-theoretic view".

Another way of answering the question (which I would coin the 
"model-theoretic view") is to ask:

are the interpretations of a rdfs:Literal included in the 
interpretations of owl:Thing?

  - in OWL DL: no!
  - in OWL Full: yes!
  - in RDF/RDFS: not determined (owl:Thing does not have a defined 
interpretation).

In informal language, when people talk about things like literals or 
owl:Things, they often refer to the things in the interpretation domain 
(the things that they denote) and not of formulas.  For instance, if I 
say "is the number 38 a rdfs:Literal?" I am not asking for the truth 
value of:

"38"^^xsd:integer rdf:type rdfs:Literal .

which, in RDF, RDFS, OWL DL and OWL Full, is a syntax error.

> In some systems this is a syntax error.
> In some systems this is necessarily true.
> In no systems is this false.

Nonetheless, I see your point: what matters in the end is what the 
system does. Note that it can be false in some systems: consider the 
following graph attached to a RDFS reasoner which supports D-entailment:

ex:g { owl:Thing rdfs:subClassOf xsd:boolean . }

With this graph, the triple rdfs:Literal rdfs:subClassOf owl:Thing is 
false. In case you consider the statement independently from any graph, 
then it is neither a syntax error, nor a true statement, in that system.


Regards,
-- 
Antoine Zimmermann
Researcher at:
Laboratoire d'InfoRmatique en Image et Systèmes d'information
Database Group
7 Avenue Jean Capelle
69621 Villeurbanne Cedex
France
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/

Received on Thursday, 14 October 2010 16:16:46 UTC