RE: OWL guide comments

Some useful comments I will try to incorporate into the Guide.

- Mike

-----Original Message-----
From: Jim Hendler [mailto:hendler@cs.umd.edu]
Sent: Wednesday, October 30, 2002 9:14 PM
To: Smith, Michael K
Cc: dder@ecs.soton.ac.uk
Subject: Fwd: OWL guide comments


Mike- Dave DeRoure ran Guide past some outsiders - got a lot of 
comments - here they are.  Dave, can we share publicly with Webont? 
- JH  (Mike is the editor - this is clearly useful and timely)


>Date: Thu, 31 Oct 2002 02:26:50 +0000 (GMT)
>From: David De Roure <dder@ecs.soton.ac.uk>
>X-Sender: dder@pandora
>To: Jim Hendler <hendler@cs.umd.edu>
>Subject: OWL guide comments
>X-ECS-MailScanner: Found to be clean
>
>Hello Jim,
>
>I mentioned I'd test the OWL guide on grid guys doing 'semantic grid'
>projects.  Here is the feedback from one of them.  Could you pass this
>onto the editor of the document please, if you feel is is useful and
>timely?
>
>Thanks
>
>-- Dave
>
>GENERAL DOCUMENT COMMENTS
>-------------------------
>
>The wine example, while entertaining and interesting, might be a little
>too clever in places and obscure the point of the examples (especially
>earlier in the document). e.g. when xmlns:vin is introduced (#Namespaces)
>it took me a good few seconds to catch on to the name; in the
>Introduction I actually didn't have any mechanism to conclude that
>WHITEHALL-LANE-PRIMAVERA (never heard of it) is a DESERT-WINE (though
>I can see that this might be the point!). Once past the introduction,
>the example does become much clearer.
>
>A glossary might be useful for some terms that might be less familiar to
>some readers.
>
>#Introduction
>-------------
>
>General: this section suffers from too many short sentences, and
>doesn't flow or read well.
>
>There is a tendency to introduce many new terms in quick succession,
>without explaining them or grounding them in terms already used (or even
>ones known/clearly understood by the reader). I guess it's unfortunate
>that some of the terms used are already overloaded in the wider field of
>computer science.
>
>e.g. The list at the end of para 3 ("an ontology may include"). In this
>context, what is a class? What is an instance? What is an element of a
>class? (is it anything to do with an instance?). While I have
>preconceptions of what these are from experience, I also know that in
>different domains they can have subtly different meanings.
>
>In the next paragraph "Datatype properties and object properties are
>collectively the properties of a class": are they properties of a class
>or an element (as alluded to in the list)? It is unclear, and makes
>harder reading than it should.
>
>Next para: "Defined classes and properties can then be used to describe
>specific Web resources": does that mean classes for a _specific_
>resource, or an instance? (although I'm not entirely clear what either
>of those are at this point ;)
>"many additional derived characteristics of the instance may": which
>instance?
>
>This para and the next really suffer from the short sentence thing.
>
>By the time we reach "There are two answers to this question" it's no
>longer a question (it's been broken up by a couple of statements).
>
>#Analogies
>----------
>
>Object oriented programming: the mapping between OWL and OOP isn't clear
>(i.e. which terms are OWL ones, and which OOP?). "Consider classes to be
>object definitions, with functional properties as fields and relational
>properties as boolean methods" sound like generic OO speak about
>nothing ;)
>
>#StructureOfOntologies
>----------------------
>
>Para 1: both "individuals" and "instance" are italicised. Are two new
>(and different) terms being introduced, or are they synonymous? This
>should be made clear.
>
>Para 4: typo "cannot it cannot"
>
>#SimpleClasses
>--------------
>Whoa, the guide jumps straight into using the terms "class" and
>"individual" without clearly defining them in this domain.
>
>#DefiningSimpleClasses
>----------------------
>taxonomy / taxonomic  ->  glossary?
>
>#DefiningIndividuals
>--------------------
>The individual / instance duality persists, although it becomes
>clearer...
>
>#SimpleProperties and #Datatypes1
>---------------------------------
>In the first of these sections, there are relations between "elements"
>and in the second between "resources" - again, would it be better to
>standardise on using one term for consistent reading (or at least
>define them in relation to each other)
>
>#OntologyMapping
>----------------
>"In the best of all possible worlds they need to be composed": composed,
>or composited? In either case, it might be better to add "from several
>existing ontologies".


-- 
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 Thursday, 31 October 2002 10:41:22 UTC