RE: Owl2DotOwlDevel comments

Hi Holger!

Since your mail contains a lot of different points, and since I need to
further think about some of them, I will answer by more than one mail. Of
course, all being said here is my own opinion, and does not necessarily
reflect the view of the OWL WG. Also, all changes that I am doing to the
proposal may be reverted by the WG later on.

>-----Original Message-----
>From: public-owl-dev-request@w3.org [mailto:public-owl-dev-
>request@w3.org] On Behalf Of Holger Knublauch
>Sent: Thursday, July 23, 2009 2:43 AM
>To: public-owl-dev@w3.org
>Subject: Owl2DotOwlDevel comments
>
>Hi Michael,
>
>many thanks for working on the owl.owl file, which I am tracking on
>the OWL mailing list and
>
>	http://www.w3.org/2007/OWL/wiki/Owl2DotOwlDevel
>
>I have copied a snapshot of the page above into an N3 file and opened
>it in TBC. It works nicely out of the box.

Good, thanks! :)

>Here are some comments:
>
>1) owl:topObjectProperty has the wrong rdf:type of rdf:ObjectProperty

No, this is really intended. And since you are tracking the discussion on
the WG list, you will find that I had proposed this already in my original
mail [1] as "exceptions", before providing the actual triples. Unlike the
other properties in the OWL 2 vocabulary, owl:topObjectProperty and
owl:bottomObjectProperty are really object property in all of the OWL 2
sub-languages that support these features, including OWL 2 DL. 

For OWL 2 Full, it does not make any difference whether to use rdf:Property
or owl:ObjectProperty as a type, but for OWL 2 DL a type of rdf:Property
would make no sense. So choosing owl:ObjectProperty seemed to me the right
way to go in this particular case. For the analog reason, I have proposed
that owl:Thing and owl:Nothing should be of type owl:Class rather than
rdfs:Class (as in owl.owl, btw., but this was not the reason to propose
this).
 
>2) I think the following classes should be rdfs:subClassOf
>owl:ObjectProperty:
>    - owl:AsymmetricProperty
>    - owl:IrreflexiveProperty
>    - owl:ReflexiveProperty
>    - owl:SymmetricProperty (this used to be in OWL 1)
>    - owl:TransitiveProperty (in OWL 1 as well)

Skipped for the moment. I have to think about it in more depth.

>3) I think it would be great to have some triples on the XSD datatypes
>in that file. I know, you may not want to have this in the OWL
>namespace, yet you are already adding triples to RDFS properties, so
>why not also for XSD?

Also skipped.

>4) The following states that owl:DataRange has been replaced with
>rdfs:Datatype
>
>http://www.w3.org/TR/2009/CR-owl2-rdf-based-semantics-
>20090611/#Vocabulary_Terms
>
>Would it make sense to add owl:DataRange rdf:type owl:DeprecatedClass ?

This would mean to make a resolution of the WG in an inherent "semantic"
aspect of this class, at least for OWL 2 Full. This would look very strange
and unmotivated to me, and technically it would change a normative aspect of
the OWL 2 RDF-Based Semantics. Since this document is a Candidate
Recommendation, it is probably too late for such a change. I have instead
added a deprecation note to the rdfs:comment of owl:DataRange, the same as
the one you have cited above. Is this ok for you?

>5) Great to see rdfs:labels and comments there! Some editorial
>comments (and sorry for those tiny details)
>
>a) I would start the comments with an uppercase letter and end them
>with a dot to form regular sentences. This will hit you back if you
>ever want to add a second sentence (separated by .). Also, most other
>ontologies seem to do that, including the RDF Schema ontology.

Ok, I have checked the rdf(s).rdf(s) ontologies, and have changed all
comments accordingly.

>b) For RDFS resources such as rdfs:label, please do not include
>comments and labels, because they are already defined in the RDFS
>namespace document
>
>	http://www.w3.org/TR/rdf-schema/#ch_appendix_rdfs

Ok, removed. They are not really needed.

>The ranges and domains of those RDFS properties are fine to have in
>the OWL document, as long as they are compatible (they will just be
>merged then). The same applies to the comments and labels - you could
>copy them if you want but please do not change them - this will cause
>confusion.

The domains and ranges are hopefully the same, otherwise it would be a bug.

>c) The comments often start with "the property for ..." such as in
>	"the property determining the cardinality of a maximum..."
>Why not just
>	"Determines the cardinality of a maximum..." ?

I have refined the comments. Please check again.

>Especially the "the" sounds strange, for example "the annotation
>property for adding a label to an entity" reads as if rdfs:label would
>be the only possible property for that purpose. Replacing "the" with
>"a" should fix that.

The comments for the RDFS annotation properties have been removed.

>d) In addition to comments, it would be great to have a reference to
>the section in the formal specifications as rdfs:seeAlso (I know,
>these URLs will change so it's probably too early now)

I have added (boring :)) rdfs:seeAlso triples to every OWL term, following
the practice in rdf(s).rdf(s).

>e) It is great to be finally rid of the "backwardCompatibleWitesh"
>typo :)

I used a regex for creating the labels. I hope that this doesn't make
owl2.owl "owl:incompatibleWith" owl.owl. ;-)

>Regards,
>Holger

Cheers,
Michael

[1] <http://lists.w3.org/Archives/Public/public-owl-wg/2009Jul/0029.html>

--
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
=======================================================================