RE: review of wine.owl and food.owl

I am about cleaned up.  I assume the Guide reference text

 Note that in order to use the OWL vocabulary you do not need to
 import the <a href="http://www.w3.org/2002/07/owl">owl.owl</a>
 ontology.  

should read

 Note that in order to use the OWL vocabulary you do not need to
 import the <a href="http://www.w3.org/2002/07/owl">owl</a>
 ontology.  

Or should "owl" be "owl.rdf"?

- Mike

-----Original Message-----
From: Jeremy Carroll [mailto:jjc@hpl.hp.com] 
Sent: Friday, July 25, 2003 7:00 AM
To: www-webont-wg@w3.org
Subject: review of wine.owl and food.owl



Mike please can you acknowledge receipt of this, otherwise I'll give you a 
call later.


Note:
due to the tardiness of this review (which makes points that I don't recall 
having previously made) I am happy for the editors to totally ignore the 
suggestions - although on the first two I would defer to DanC and Sandro ...

 
1: Base location

It is an improvement that xml:base is used.
I think it is probably more appropriate to use the base location that
includes 
the publication URI so e.g. for the current WD it would be

http://www.w3.org/TR/2003/WD-owl-guide-20030331/wine#

and say the CR gets published 10 August it maybe

http://www.w3.org/TR/2003/CR-owl-guide-20030810/wine#

This is a bit of pain at publication time, but is probably better than the 
implicit link off to the changing location

 "http://www.w3.org/2001/sw/WebOnt/guide-src/wine#"

2: Changing suffix to .rdf

Given that we decided not to register a mime type I think we should be using

the .rdf suffix rather than a .owl suffix.

3: (if doing 1)
If you decide to follow the suggestion in point 1, then it may be helpful to

replace the namespace declarations with entity refs to minimize the points 
of change.

e.g.
<rdf:RDF
  xmlns     = "&vin;"
  xmlns:vin = "&vin;"
  xml:base  = "&vin;"
  xmlns:food= "&food;"
  xmlns:owl = "&owl;"
  xmlns:rdf = "http://www.w3.org/1999/02/22-rdf-syntax-ns#"
  xmlns:rdfs= "http://www.w3.org/2000/01/rdf-schema#"
  >


4: making food.owl conform with OWL DL
(I can't remember if we have already discussed this - if you have already 
rejected this suggestion apologies for repeating myself)

food owl is not in OWL DL because it refers to the objects in the wine 
ontology without giving them types e.g.
  <owl:Class rdf:ID="Wine">
    <owl:equivalentClass rdf:resource="&vin;Wine"/>
  </owl:Class>

This can be rectified in three different ways:

a) easy way
    import wine.owl from food.owl

b) harder way
    include specifc type information in food.owl, e.g. above fragment
becomes

  <owl:Class rdf:ID="Wine">
    <owl:equivalentClass>
         <owl:Class rdf:about="&vin;Wine"/>
    </owl:equivalentClass>
  </owl:Class>

c) very hard way
   make a new file which just has the type information for both wine.owl and

food.owl and import them.

5: delete xmlns:xsd="...." from wine.owl

This is not used 

6. Keeping wine.owl in OWL DL - imports object

Food.owl declares its own URI to be of type owl:Ontology using the xml:base 
and the idiom
   <owl:Ontology rdf:about="">

The xml:base ends in a # which gets ignored when resolving the empty same 
document reference, so that the subject of the rdf:type owl:Ontology triple 
is the URIref
http://www.w3.org/2001/sw/WebOnt/guide-src/food

However the object of the owl:imports triple is

http://www.w3.org/2001/sw/WebOnt/guide-src/food.owl

While the imports mechanism works, this uriref remains untyped and so the 
document is in OWL Full.

Deleting the .owl suffix may work (it definitely will work if you use .rdf
as 
your suffix). 
Including the suffix in both places will also work.

(I am still working on this review - there may be one or two more points)
 
Jeremy

Received on Thursday, 31 July 2003 12:21:15 UTC