Re: OWL Web Ontology Language Reference Editorial Comments

Lacy . Lee wrote:

 > The OWL Reference document is very solid and helpful.


Thanks.

 > The only general comment I have is that the reference document
 > occasionally refers to "triples".  It the document consistently referred
 > to axioms and statements, many readers would not have to apply RDF
 > knowledge of triples to understand the content of the document.  There
 > are very few mentions of "triples" (e.g., 3.1.2, 7.1, 8.3), so I think
 > this would be a minor fix.


I prefer not to make this change. The document assumes some basic 
knowledge of RDF (see the intro).

 > I would like to offer a few nit picky comments to support the final
 > editing review:
 >
 > Section 3, NOTE beginning "If one provides", typo, change "enumerattion"
 > to "enumeration"


Fixed.

 > Section 3.2.1, first paragraph, typo, change "than" to "then"


Fixed.

 > Section 3.2.1, last paragraph, recommend changing "has a librettist" to
 > "has at least one librettist".  The "a" suggests a cardinality of 1,
 > instead of a mincardinality of 1.


Fixed.

 > Section 3.2.2, NOTE beginning "OWL DL", appears to suggest that in OWL
 > Lite a property restriction could be the subject of an
 > owl:equivalentClass statement, and I thought it could only be the
 > object.


You're right. It is stated correctly in Sec. 8,3 but we forgot to update 
the text  in this section. The last sentence of the note now reads:

[[
In OWL Lite the subject must be a class name and the object must be 
either a class name or a property restriction.
]]

 > Section 3.2.3, paragraph beginning "This class axiom", I believe you
 > meant to say "second example" instead of "first example"


We meant the first *opera* example. Changed accordingly.

 > Section 3.2.4, Man example, I believe a "Man" should be "#Man" since
 > it's an "about" attribute


Fixed.

 > Section 3.2.4, MusicDrama example, I believe a "MusicDrama" should be
 > "#MusicDrama" since it's an "about" attribute


Fixed.

 > Section 4, paragraph beginning "An object property are" should be "An
 > object property is"


Fixed.

 > Section 4.1.3, last NOTE, suggests a singular use of rdfs:range, while
 > multiple statements are allowed, recommend changing wording to
 > "...descriptions allowed as objects of rdfs:range are class names"


Changed as suggested.

 > Section 4.3.1, paragraph starting "The following", change "person" to
 > "woman"


Changed as suggested.

 > Section 5.1, paragraph starting "The example", change "with as datatype
 > the" to "with the"


Changed as suggested.

 > Section 7.1, Annotation properties aren't very well defined in any of
 > the OWL documents.


This was observed by a number of people. The editors of the OWL 
Reference have an action to add some explanatory text about this issue.

 > Section 7.1., MusicDrama example, I believe a "MusicDrama" should be
 > "#MusicDrama" since it's an "about" attribute


Fixed.

 > Section 7.2, type, change "so may" to "so it may"


Fixed.

 > Section 7.4.3, seems to be inconsistent with section 6 of the OWL Guide
 > because the Guide says not to assume backward compatibility and this
 > section seems to suggest that compatibility can be assumed for 2
 > versions

Is it posible that you misread this?
In Sec. 7.4.3 it says:

[[
If owl:backwardCompatibleWith is not declared for two versions, then
compatibility should not be assumed.
]]


 > Section 7.4.5, it's unclear in the example whether we need to explicitly
 > state that Car is an OWL class since DeprecatedClass is a subclass of
 > rdfs:Class and not owl:Class; similarly for the hasDriver property, do
 > we need to explicitly state as an owl:ObjectProperty since
 > DeprecatedProperty is a subclass of rdf:Property?

The typing triples are only needed if the ontology is explitly stated to 
be in OWL DL. That is not the case here.

 > Appendix A, versionItem should link to section 6 of the OWL Guide

Link added.

 > Appendix B, recommend providing versionInfo as a typedLiteral using
 > &xsd#date datatype in Ontology header

At the moment the value does not conform to xsd:date, so we left this 
unchanged.

 > Appendix B, shouldn't we explicitly state that the inverseOf property is
 > symmetric?


We have deliberately refrained from using OWL constructs in this schema.

 > Appendix C, isn't the domain and range of versionInfo owl:Thing?
 > Nothing is stated in the table


This is on purpose. There are no dmain/range restrictions for VersionInfo.

 > Appendix D, aren't OntologyProperty, AnnotationProperty, and DataRange
 > new since DAML+OIL?


Added these to the list.

 > I hope these comments/suggestions help!


Yes, a lot. As editor you tend to get blind at some point for errors. 
The work you put into this detailed review is very much appreciated.

You can check the changes in the editor's draft at:

   http://www.cs.vu.nl/~guus/public/owl-ref-recdraft.html

(temporarily NOT linked from the WG home page).

Guus

 >
 > Lee



-- 
Free University Amsterdam, Computer Science
De Boelelaan 1081a, 1081 HV Amsterdam, The Netherlands
Tel: +31 20 444 7739/7718
E-mail: schreiber@cs.vu.nl
Home page: http://www.cs.vu.nl/~guus/

Received on Friday, 23 January 2004 11:59:07 UTC