W3C home > Mailing lists > Public > www-webont-wg@w3.org > December 2003

Re: Annotation properties with range and other rdf:types

From: Jeremy Carroll <jjc@hplb.hpl.hp.com>
Date: Tue, 16 Dec 2003 11:35:24 +0100
To: <www-webont-wg@w3.org>
Message-ID: <BHEGLCKMOHGLGNOKPGHDEEHECCAA.jjc@hpl.hp.com>


[[I would value team contact advice as to whether this is appropriate
discussion of comments from the comment list at this point, i.e. proposing
textual changes in the light of comments - or are we now into proposing
errata in light of comments]]


Looking at thread starting
http://lists.w3.org/Archives/Public/public-webont-comments/2003Dec/0000

Jim is clearly correct to say either you use DL and respect the constraints,
or you use Full and are unconstrained.

However, I wonder whether we should have gone one step further in our last
discussion of annotation properties when we decided
"URI's used as object of annotation property to not need to be typed."
http://lists.w3.org/Archives/Public/www-webont-wg/2003Oct/0176

In particular I wonder whether the text in S&AS that says:

http://www.w3.org/TR/2003/PR-owl-semantics-20031215/mapping.html#4.2
[[
3. the ontologies in O only use the class-only vocabulary as class IDs; only
use the datatype-only vocabulary as datatype IDs; only use rdfs:Literal in
data ranges; only use the property-only vocabulary as datavaluedProperty
IDs, individualvaluedProperty IDs, or annotationProperty IDs; and do not
mention any disallowed vocabulary.
]]

should be clarified? or weakened? to

[[
3. the ontologies in O, ***except as the values of annotations,*** only use
the class-only vocabulary as class IDs; only use the datatype-only
vocabulary as datatype IDs; only use rdfs:Literal in data ranges; only use
the property-only vocabulary as datavaluedProperty IDs,
individualvaluedProperty IDs, or annotationProperty IDs; and do not mention
any disallowed vocabulary.
]]

(It is currently not clear which, if any, parts of this constraint apply to
the values of annotations - other parts of S&AS and the test document are
clear that this constraint does not prevent classIDs from being the objects
of annotations - it seems pushing it to read "do not mention any disallowed
vocabulary" as "do not mention any disallowed vocabulary except as the
values of annotations")

My understanding is that this clarifies the intent.

So if I were to draft a response to Holger, with this change it could be:

[[
To remain within OWL DL, and to declare a "range" on annotation properties
for the purposes of a GUI or similar, you should use your own range
property.
e.g.
<owl:AnnotationProperty rdf:about="&my;sameAs">
  <rdfs:comment>This is informally like owl:sameAs,
   intended for documenting AnnotationProperties which in turn
   can be used for documenting AnnotationProperties</rdfs:comment>
  <my:sameAs rdf:resource="&owl;sameAs"/>
</owl:AnnotationProperty>
<owl:AnnotationProperty rdf:about="&my;range">
  <rdfs:comment>This is informally like rdfs:range,
   intended for documenting AnnotationPropertys</rdfs:comment>
  <my:sameAs rdf:resource="&rdfs;range"/>
</owl:AnnotationProperty>
<owl:AnnotationProperty rdf:ID="intValueAP">
  <my:range rdfs:resource="&xsd;int"/>
</owl:AnnotationProperty>


The informal constraints on annotation properties can be made formally by
the addition of the single triple
my:sameAs owl:sameAs owl:sameAs .
which, takes the ontology into OWL Full
]]

Jeremy
Received on Tuesday, 16 December 2003 05:36:15 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:58:03 GMT