|
Pat,
I'm delighted to see you raise this matter
again!
And to show you just how delighted I am, let me
retrieve an introductory piece of a post I was working on just before I had the
idea of taking part in Steve Ray's lead-up to the Summit on March
26.
I had been explaining, in my unsent draft
post, why my "5th instalment" had been so long delayed but why, together
with my very premature announcement of my website on TheMainstream.info, I was
nonetheless venturing out. This is how I had put it at that
time:
My tactical
move, however premature, was triggered by two recent or current ontolog-based
activities:
(1) The apparent and
regrettable stalemate ending the "Standard Ontology" thread, between
Pat Cassidy with his FO concept (Foundation Ontology, though he has also used
"Common Foundation Ontology") and almost all other participants. That
thread, beginning in earnest with Pat C's post now at http://ontolog.cim3.net/forum/ontolog-forum/2009-02/msg00138.html,
had arisen out of this recognition of a major opportunity:
(2) OntologySummit2009, with
its challenging theme, "Toward Ontology-based Standards", as proposed by
Steve Ray at http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2009.
The Summit’s looming date a bare
few weeks ahead draws me to indicate how we might together resolve - cut?
untie? - the FO Gordian knot before it tightens or is brushed
aside.
So now I can point out that the
"Summit" slideshow I introduced in my recent post now at http://ontolog.cim3.net/forum/ontolog-forum/2009-04/msg00016.html does
allude briefly to that part of my proposal that addresses the
ontology-for-universal-interoperability requirement you have been approaching
from another angle.
Today's plain but downloadable
version of the slideshow is at http://TheMainstream.info/RideTheMainstream20090406.pps,
the annotated but online version at http://TheMainstream.info/RTM.html.
Slide 30, bullet 2, is the allusion
itself. But the whole slideshow with its now-added Notes (only in the
.html version), together with its references on its slide 18 to earlier
instalments of my "MACK basics" series from last year, does build up to
that bullet, which might even be identified with the core "Silver
Bullet" that my above-linked post went on so long about!
So do please try extra hard to
understand that slideshow and its notes (which I admit is not an easy
matter...). And do please ask for clarifications where necessary, and make
any comments that might come to mind.
Paola, the same applies to you, with
your question below: "Dont we need a kind of
'meta ontological framework'?"
Yes we do, and it's implicit in the
subject of that slideshow.
Christopher
----- Original Message -----
Sent: Monday, April 06, 2009 3:54
AM
Subject: Re: [ontolog-forum] meta
ontology framework
Paola,
At the risk of resurrecting that long discussion, I don’t
think this should pass:
[PD] > But how, I asked , how are we going to get
all these little ontologies to work with one another and possibly synchronise?
Dont we need a kind of 'meta ontological framework'? yes,
said Fabio, everyone who is delivered an ontology needs a framework for it to
be casted in and to align other ontologies with. And who is doing such work?
Nobody that anybody knows of.
Well, do recall the proposal I have been discussing for a Foundation Ontology
as the basis for logical descriptions of any domain ontology that its creator
actually wants to interoperate with others. That ontology, and methods
for extracting parts for use in domain ontologies, is, in my view,
precisely the kind of ‘metaontological framework’ that would enable accurate
interoperability. My own efforts, very limited compared to what
needs to be done by a large consortium, are aimed at providing evidence that
such an approach is technically sound. It is not true that *no
one* is working on the issue, but it is true that a lot of people are
ignoring the issue. There has also been a lot of effort to find methods
to map ontologies to each other, which I think may be useful in some limited
contexts. But I cannot visualize any way that automated mapping can
create alignments accurate enough for mission-critical applications.
Semiautomated mappings could help, but unless they are mapped to a common
foundation ontology that has sufficient content to support translation among
alternative representations, the cost of one-on-one semiautomated alignment
will be probative in all but the most pressing cases.
The
cost of delay in initiating a consortium effort to develop an Foundation
Ontology acceptable to a large community (it doesn’t need everyone to agree)
is very high, both in lost efficiency of noninteroperable commercial
applications, losses due to hopeless and unnecessary pursuit of alternatives,
and inefficiency in research on AI generally because of the absence of a
common knowledge representation that will allow more efficient reuse of
research results. By comparison, the cost of a serious effort to
directly address the problem with a common Foundation Ontology is
miniscule. Be skeptical if it suits you, but please don’t say that no
one is making an effort to address the problem directly.
Pat
Patrick
Cassidy
MICRA,
Inc.
908-561-3416
cell:
908-565-4053
cassidy@xxxxxxxxx
From:
ontolog-forum-bounces@xxxxxxxxxxxxxxxx
[mailto:ontolog-forum-bounces@xxxxxxxxxxxxxxxx] On Behalf Of
paola.dimaio@xxxxxxxxx
Sent: Sunday, April 05, 2009 1:47
PM
To: [ontolog-forum]
Subject: [ontolog-forum] meta
ontology framework
an earlier excerpt from a previous post is going around my
head (sunday blues)
CM
The puzzle is why you don't
grasp the fact that the writer is
obviously arguing against the idea of a
single ontology.
At the cost of getting scolded, this discussion is
far from exhausted.
I had this conversation with Fabio Ciravegna
after a keynote speech, where he talks intesively of the small ontology
world (federated, distributed, networked ontology). Point taken.
Agreed,
But how, I asked , how are we going to get all these little
ontologies to work with one another and possibly synchronise? Dont we need a
kind of 'meta ontological framework'? yes, said Fabio,
everyone who is delivered an ontology needs a framework for it to be casted in
and to align other ontologies with. And who is doing such work? Nobody that
anybody knows of.
Having read relevant portions of Azamat's work, I
think what he calls unified ontology, and therefore raises eyebrow and
argument like the ones we have had on thi list, and the one above, is a meta
ontological framework (not an ontlogy metamodel like the OMG, but in the same
direction). A system to align and support the synchronization of different
part-ontologies
I have no doubt that this kind of work is needed,
although I think maybe it does not always come across cleary what is it
exactly, what purpose, what benefits
I think avoiding to deliver on a
unified framework, means that
1. the (domain, application, task)
ontologies alone will never be able to be integrated seamslessy and
dynamically aligned with other ontologies and its overall information
environments
2. the gaps created by such built in systemic ontological
misalignment will be very expensive to fix, and the 'need for more research'
cycle will be self perpetrating (and oh yes, a consortium of top universities
will deliver that..)
I encourage or anyone who intends to bring a
novel perspective (and is trying to get funded) to
make sure that their
proposed ideas are demonstrable and directly aimed at fill existing
gaps.
Now burn me.
pdm
--
Paola Di Maio,
****************************************
Forthcoming
IEEE/DEST 09
Collective Intelligence Track (deadline extended)
i-Semantics 2009, 2 -
4 September 2009, Graz, Austria. www.i-semantics.tugraz.at
SEMAPRO
2009, Malta
http://www.iaria.org/conferences2009/CfPSEMAPRO09.html
**************************************************
Mae
Fah Luang Child Protection Project, Chiang Rai
Thailand
_________________________________________________________________
Message
Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config
Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files:
http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J
To
Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx
|