Annotation properties - a thought

Another email has shown up on our public-comments-list asking about 
annotations.  I think part of what is going on here is that 
annotations were a relatively late addition to OWL, and thus didn't 
get integrated into our documents as much as some other features did. 
I would NOT advocate changing documents at this point, but wonder if 
someone would like to generate a short discussion of annotation 
properties -- what they are for, how they are used, how they can be 
used in DL(i.e. can I create a subclass of annotation property and 
still be in DL). how they can be used at all (can I make an 
annotation property that is an object type in Full)?
   If we were to have such a thing somewhere in our mail archive, or 
better or web site, it would be useful for now -- and if/when the 
Best Practices WG starts, this could perhaps become part of an OWL FAQ
  Just a thought - would be great if someone would volunteer to do this.
   -JH
p.s. would be useful if someone could answer the message below which 
is copied from 
http://lists.w3.org/Archives/Public/public-webont-comments/2004Jan/0006.html


At 15:49 +0100 1/9/04, Benjamin Nowack wrote:
>
>Hi,
>
>it would be great if someone could confirm/answer the following,
>I'm not 100% sure if I got it right. The questions are partly related
>to Holger Knublauch's comments on annotation properties in OWL DL [1].
>
>1) I would like to use dc:creator as a DatatypeProperty. If I say
>    <owl:DatatypeProperty rdf:about="&dc;creator" />, am I still
>    in OWL DL space?
>
>2) I would like to use dc:creator to annotate my ontology. Now
>    I must not have the above statement in my ontology if I don't
>    want it to become OWL Full. Right?
>
>3) <owl:AnnotationProperty rdf:about="&dc;creator"/> allows me to
>    use it on classes, properties, individuals and ontology headers,
>    but reasoners will ignore them. Does that mean, that API builders
>    should skip annotation properties in e.g. rdql query patterns? Or
>    does this only mean that annotation properties should not be used
>    for class modeling etc.?
>
>
>I appreciate any help. Thanks in advance.
>Benjamin
>
>
>[1] 
>http://lists.w3.org/Archives/Public/public-webont-comments/2003Dec/0000.html



-- 
Professor James Hendler			  http://www.cs.umd.edu/users/hendler 
Director, Semantic Web and Agent Technologies	  301-405-2696
Maryland Information and Network Dynamics Lab.	  301-405-6707 (Fax)
Univ of Maryland, College Park, MD 20742	  240-277-3388 (Cell)