Re: review of Guide

From: Christopher Welty <welty@us.ibm.com>
Subject: Re: review of Guide
Date: Wed, 1 Jan 2003 10:52:09 -0500

> 
> Peter,
> 
> In the guide we wanted to make a distinction between what many people call 
> "an ontology" and what an ontology is in the OWL language.  In OWL, 
> ontology is just a syntactic construct - I don't believe it has any 
> semantics, and there is nothing about the tag itself that enables anyone 
> or any system from distinguishing what is conventionally called an 
> ontology from what can be in a document taged as one.
> 
> I am fairly abmivalent about this issue - Pat argues vehemently on both 
> sides, having yelled at me several times for trying to right the great 
> wrong that's been done to the word "ontology" by computer scientists (who 
> are deeply confused about what it actually did mean), and then yelling at 
> anyone else who tries to argue for the very distinction computer 
> scientists claim the word describes.
> 
> Anyway, there is an established and growing community in computer science 
> (and, now, linguistics and medicine, too) who want to distinguish a 
> "collection of classes and relations" from that collection plus all the 
> instances of those classes and relations.  They call the former an 
> ontology.  They call the latter any number of things, including a 
> knowledge base.  OWL, syntactically and semantically, has no such 
> distinction( and furthermore with OWL Full makes any collection of 
> instances a potential "ontology" anyway). But the group wanted to make the 
> point, and I agree it needs to be made.
> 
> Whether or not the term "knowledge base" is an "official" OWL term, I want 
> to use it and "ontology" in Guide (I admit it needs to be done more 
> rigorously) to refer to these two distinct things.  There is some text I 
> added that mentions that we use owl:ontology to talk about OWL documents, 
> and ontology and knowledge base FOR THE PURPOSES OF THE GUIDE ONLY to talk 
> about the other two kinds of things.
> 
> Happy new year to all.
> 
> -Chris

If you want to make this distinction in the Guide, then you are going to
have to come up with a distinction that you can define and communicate this
distinction to the rest of the working group.  The other documents will
probably all then have to be rewritten to conform to this distinction.  As
it is, the Guide now introduces an important concept that it not mirrored
in the other documents.

peter

Received on Wednesday, 1 January 2003 11:44:53 UTC