GUIDE: review of (9/18) GUIDE document

Mike - as a way to review the OWL in the guide document, I decided to 
use it to explain OWL to my graduate semantic web class (explaining 
to them we were reviewing the document which was very much in 
progress).  Short summary is we found tons of syntax errors, but 
overall found it a useful document and an improvement on the previous 
walkthru in some ways.

This review was of the September 18th version [1], apologies if there 
is one later than this.


Meta comments-
  History - problematic as is - Peter's changes, my changes to his 
changes, and Leo's changes to my changes to his changes should all be 
mentioned (OML doesn't actually need to be included - it never 
actually got implemented, but XOL did)

Wine thread -
  the students and I all liked the fact the example ran through the 
document, However, those cases where it didn't (property 
characteristics and restrictions, for example) were somewhat jarring 
- probably be better to find examples for those we can (not that 
important, but improves readability).

owl:individual -
  With due respect, there's no such thing -- in the features document 
individual is there as a descriptive term - as I understand OWL, we 
do not say:

><owl:Individual ID="CENTRAL-COAST">
>   <type rdf:resource="#CALIFORNIA-REGION"/>
>  </owl:Individual>

but rather
  <CALIFORNIA-REGION rdf:id="CENTRAL-COAST" />

(and when the definition contains other properties it looks like)

     <CAL-REGION rdf:about="http://foo.bar.baz#CAL-COAST">
         <cal-location>WEST</cal-location>
     </CAL-REGION>

(see [2] for other, more complex examples of an individual.)

Consistency
  there's a lot of inconsistency in capitalization (owl:ontology and 
owl:Ontology for example) and also some typos like:

>   <owl:Class rdf:ID="CABERNET-SAUVIGNON">
>   <rdfs:subClassOf>
>     <rdfs:subClassOf rdf:about="#WINE"/>
>   </rdfs:subClassOf>

which is incorrect - you don't need the extra subclass level.
   When this document is further along, it will need a VERY thorough 
scrub of the RDF/OWL (Ora Lassila is great at this - maybe we could 
recruit him?)

OneOf
  I think the use of owl:thing in your OneOf example is legal, but I 
think the preferred form is to reflect the property name -- that is, 
in the DAML walkthru the example

><daml:Class rdf:ID="Height">
>   <daml:oneOf rdf:parseType="daml:collection">
>     <Height rdf:ID="short"/>
>     <Height rdf:ID="medium"/>
>     <Height rdf:ID="tall"/>
>   </daml:oneOf>
></daml:Class>

is the preferred way.

Again, sorry to pick so many nits - overall the document is a good 
contrbution, shows off the key language features and has a better mix 
of instances and classes than the DAML walkthru did.




[1] 
http://lists.w3.org/Archives/Public/www-webont-wg/2002Sep/att-0257/01-Guide.html
[2] http://www.cs.umd.edu/~hendler/jhendler.daml

-- 
Professor James Hendler				  hendler@cs.umd.edu
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	  240-731-3822 (Cell)
http://www.cs.umd.edu/users/hendler

Received on Tuesday, 24 September 2002 21:57:32 UTC