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

Review of Guide

From: Jim Hendler <hendler@cs.umd.edu>
Date: Wed, 1 Jan 2003 18:49:23 -0500
Message-Id: <p05200f4fba391f5ad25f@[10.0.1.3]>
To: webont <www-webont-wg@w3.org>

A review of http://www.w3.org/2001/sw/WebOnt/guide-src/Guide.html - 
18 december draft.

Abstract - perhaps change order of 1st and second paragraph so that 
OWL is mentioned up top (seems more typical)

[Note: use of term knowledge base (in ref to Peter's earlier remarks) 
- Ref defines KB consistent with use in Guide, so I'm comfortable 
with current usage in Guide]

Meta-comment, this document is designed for use with OWL DL.  This is 
okay, but that should be made clearer up top, and it might be worth 
considering an appendix which gives examples of Full, as has been 
suggested by some of us previously.



1 - intro, why "and to a lesser degree" instances?  There are 
examples in some of our libraries w/small number of classes but lots 
of instances all in one doc...  in general this whole document tends 
to be a little "perjorative" of instances, most of which I can live 
with, but this one seems unecessary, so I would either remove the 
"and to a lesser degree" or replace it with "as well as"

2 - this issue also leads into awkwardness in the next paragraph which says

>The owl:Ontology tag does not necessarily identify a document as an ontology.

which probably should be rewritten to specify more clearly what it is 
trying to imply:

The owl:Ontology tag. therefore, does not necessarily identify a 
document as an ontology containing only class definitions with no 
instances.

3) end of Species of OWL reads:

>When we introduce constructs that are only permitted in OWL DL or 
>OWL Full, they are marked by "[OWL DL]".

this is ambiguous - do you mean that since most constructs are in 
both (i.e. the vocabulary is the same) you will only use [OWL DL} to 
mean it is in both, or do you mean you'll use [OWL DL] and [OWL Full] 
respectively?  I think you should rewrite to make clear.

4) Political rather than technical point.  The 3 paragraphs at the 
start of "Structure of Ontologies" section seem rather severe - 
rather than "come use these nice toys" they have a "careful or your 
hands will be burned" flavor - I don't think these should be deleted, 
but perhaps they can be moved later in the document - I'm not sure 
they're really needed here - perhaps go right to the namespaces 
stuff, and put these paragraphs as "caveats" somewhere later.


5) the section Defining Simple Hierarchical Named Classes has a link 
on the word "extend" in the 4th paragraph - but following the link 
doesn't seem to get me to anything about extending things - is this 
link intended?

6) datatyping stuff will need to be used carefully in this document 
which will require extra care -- I suspect you'll need to add a 
discussion at the beginning of the properties and datatypes section


7) fix typo in "inversefunctionalProperty" section -
>In OWL Full, we can use tag (sic) a DatatypeProperty as 
>inverseFunctional. This permits us to identify a string as a unique 
>key.

also, why not include an example of this, with one of your little red 
thingies - since it will be indicated (and not in your ontology), 
there's no reason you couldn't include it (even though it is in Full).


8) the example of some and all values from could be improved, because 
this is an important function of OWL and one that is hard to 
understand. saying a wine has all makers or some makers is a bit odd 
and thus it makes the example less intuitive.  Not sure what to 
suggest, maybe one of the wine mavens can come up with something more 
intuitive - my own best examples of all v. some come from the domain 
of genetics (i.e. all diseases efdecting this location  are cancers 
v. at least one of the diseases is a cancer), so they don't map to 
this domain so well ...

9) cardinality will need to be fixed in wineYear example.  Also, 
maybe put in a pointer to the "CardinalitySyntaxNote" in the 
reference document, which has an alternate syntax that is sometimes 
desirable.

10) need to add allDifferentFrom after differentFrom if the WG approves this


11) I don't have the energy to check the cross reference table - but 
it is very valuable, so let's make sure someone goes through that at 
some point to make sure pointers are correct.

12) References -"related OWL" - the list is out of date and some 
pointers wrong - make sure to fix.

13) very minor - perhaps we should alphabetize the references in the 
various reference section - wil make it easier to find/check at a 
later date

14) appendix B - history - status of RDF S has changed - make sure we 
get it right

15) I'm still concerned about mentioning particular users/usages in 
appendix B (lots of other compnanies are now using DAML or OWL, and 
so are other Gene developers, software folks, etc) - mayve we should 
cut this section after the paragraph about 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 Wednesday, 1 January 2003 18:49:27 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:57:56 GMT