Re: Good and Bad IS-A hierarchies

hovy@isi.edu (Eduard Hovy)
Date: Wed, 7 Jun 1995 16:45:19 -0700
X-Sender: hovy@quark.isi.edu
Message-id: <abfb7ac75f0210046f53@[128.9.208.191]>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
To: srkb@cs.umbc.edu
From: hovy@isi.edu (Eduard Hovy)
Subject: Re: Good and Bad IS-A hierarchies
Cc: cg@cs.umn.edu, fritz@rodin.wustl.edu
Sender: owner-srkb@cs.umbc.edu
Precedence: bulk

At 10:01 PM 6/7/95, Fritz Lehmann wrote:
>     Peter Clark wrote, among other things, this:
>-----begin quote----
>
>What I wouldn't like to
>now see is a lot of isa-hierarchies published, with little or no information
>in the general/top-level concepts.
>----end quote----
>
>Fritz's comment: 
>     The key thing is the "little or no information".  I _would_ like
>to see generally  useful IS-A hierarchies published, provided they
>explicity state what information is contained in each concept --
>what can be inherited.  Also there is a lot of important Large-scale
>structure stuff that normally goes unaddressed, like the different
>notions of "dimension" in Bruce Porters, Doug Skuce's, CYC's (slot-
>based still?) and my theories.  I don't know how the Pangloss
>Ontology handles this; I've never seen the ontology itself.
>
>     Peter Clark wrote, among other things, this:
>----begin quote----
>... For example, your ontology may model a room as a container but I want to
>model it as a point location (or person as a thing/process, carpet as
>in/partof the room etc.). This ontological distinction isn't really
>important (it's a task-dependent question), and shouldn't distract the
>debate -- what is important is that I have the components available in
>the first place (ie. a representation of what it *means* to be a "container"
>or a "point location") to be able to build either of these models as needed.
>----end quote----


Things are generally well-behaved; if I model dimensions one way and you 
want to model them another, then when you understand my system you can 
fairly easily convert it to yours (often simply using emacs macros).  The 
issue is getting enough terms and relations in some (any!) systematic 
organization in the first place.  

So I disagree with Peter Clark.  I *would* like to see a lot of concept 
organizations (isa hierarchies at the very simplest), presented in a way 
that is easily browsable so that I can pick and choose bits and pieces 
from here and there and weld them together to suit my needs of the moment.  
(Or is anyone going to say they know the true, correct, way to model the 
world??) 

E

----------------------------------------------------------------------------
Eduard Hovy
email: hovy@isi.edu          USC Information Sciences Institute 
tel: 310-822-1511 ext 731    4676 Admiralty Way 
fax: 310-823-6714            Marina del Rey, CA 90292-6695 
project homepage: http://www.isi.edu/natural-language/nlp-at-isi.html