RE: OWL Full Features in QRG

Jie Bao wrote:

>There is a side issue not related to QRG: if the use of
>"owl:OntologyProperty" is only an OWL 2 Full feature, does that mean
>all OWL 1 DL ontologies in the RDF syntax that use
>"owl:OntologyProperty" or a non-predefined "owl:OntologyProperty"
>(predefined ones will be annotation property) will be OWL 2 Full
>ontologies?

This can only be answered by the reverse RDF mapping, so let's see! 
The reverse RDF mapping refers to "owl:OntologyProperty" in three places:

* In Table 4, if I correctly understand the condition there, 
it is said that for the given ontology IRI o, there must not be any 
triple of the form 

  "o2 someOntologyProperty o"

in the to-be-mapped RDF graph. I think this is not relevant for our 
discussion, since it seems to only try to avoid that a single RDF 
graph effectively contains /two/ or more distinct ontologies.

* Table 5 performs some clean up: I there is a pair of triples

  x rdf:type owl:OntologyProperty .
  x rdf:type rdf:Property . 

then the "rdf:Property" triple is filtered out. Not problem here,
since the "owl:OntologyProperty" triple is still in.

* Finally, in Table 6, if there is a triple of the form

  *:x rdf:type owl:OntologyProperty .

then this triple will be replaced by

  *:x rdf:type owl:AnnotationProperty .

Now this is the interesting bit: This parsing rule tells that if there
is some *custom* ontology property, it will be treated just like a
custom *annotation* property by the mapping. And OWL 2 DL, of course,
supports custom annotations well.

So it looks to me that there is no danger to fall into OWL Full just for
the reason that there are ontology properties used in an RDF graph. 
On the one hand, the pre-defined ontology properties are treated by
the reverse mapping anyway in some form: 

  * either by having explicit treatment for them, as in the 
    case of "owl:imports";

  * or by treating them as pre-defined annotation properties,
    as defined in §3.5 of [Structural Spec] (somewhat hidden,
    see below for further details). 
  
On the other hand, custom ontology properties are simply treated like 
newly defined annotation properties, so the situation seems fine, too.

(Btw, I would consider it bad practice even in OWL Full to define 
new ontology properties; I consider them to be a built-in feature.)

APPENDIX: Here's how the pre-defined annotation properties of OWL 2
(including the ontology properties that have no specific treatment by the 
reverse mapping) get into the reverse mapping process:

    1) Table 9 in [RDF Mapping] initializes the annotation function "AP"
       by all annotation declarations already in "AllDecl(G)".
       The function AP is then used in the ongoing parsing process
       to represent the annotation (aka ontology) properties.

    2) The set "AllDecl(G)" is defined in §3.6 of [Structural Spec], 
       in Step CP 3.1 of the canonical parsing process. The set is 
       filled from the declarations in Table 5 of [Structural Spec].

    3) Table 5 of [Structural Spec] doesn't yet list the annotation
       properties, but rather refers to Section 5.5 for the 
       pre-defined annotation properties of OWL 2.

    4) Section 5.5 of [Structural Spec] finally lists the ontology 
       properties as annotation properties.

Michael

--
Dipl.-Inform. Michael Schneider
Research Scientist, Dept. Information Process Engineering (IPE)
Tel  : +49-721-9654-726
Fax  : +49-721-9654-727
Email: michael.schneider@fzi.de
WWW  : http://www.fzi.de/michael.schneider
=======================================================================
FZI Forschungszentrum Informatik an der Universität Karlsruhe
Haid-und-Neu-Str. 10-14, D-76131 Karlsruhe
Tel.: +49-721-9654-0, Fax: +49-721-9654-959
Stiftung des bürgerlichen Rechts, Az 14-0563.1, RP Karlsruhe
Vorstand: Prof. Dr.-Ing. Rüdiger Dillmann, Dipl. Wi.-Ing. Michael Flor,
Prof. Dr. Dr. h.c. Wolffried Stucky, Prof. Dr. Rudi Studer
Vorsitzender des Kuratoriums: Ministerialdirigent Günther Leßnerkraus
=======================================================================

Received on Thursday, 21 May 2009 21:25:26 UTC