Ed,
Just to comment on one of your points:
[EB >
> Now, if John can proffer approaches for judging the quality,
> extensibility, utility, etc., of a "fundamental ontology" (as Chris
> Welty tried to do a few years back), THAT would be (IMO) of direct
> value to the study group.
>
Based on the many discussions of ontologies over the past fifteen years,
I think that there is no good method for judging the quality of an ontology
that does not include observation of how it operates in practice in
supporting applications. Ontologies are too complicated to predict their
behavior from an inspection of individual components. In this respect, they
are a lot like software code. As I see it, the absence of publicly
manipulable applications for ontologies that support reasoning is the main
barrier to wide adoption, and an insuperable barrier to broad agreement on
any foundation ontology. Developing such applications should be the top
priority at this point. It is utterly inadequate for proponents of any
ontology to cite uses that cannot be observed and tested first-hand by other
potential users. No one can determine whether an ontology is in fact useful
for its intended purposes without being able to test for oneself how it
operates; passively watching a canned demo tells one very little, and
hearing second-hand stories about who has done what with an ontology has no
value; we need the ability to test an ontology-based application and
determine for ourselves its strengths and weaknesses. It is very
disappointing to me that at this late date in the development of ontology
technology there are still only publicly available ontology *development*
environments such as Protégé and SWOOP and their accessories (important, and
excellent in many ways) - But no ontology-driven *practical* applications
that we can observe in operation and test for ourselves.
I do not include as "applications" the use of an ontology as a more
formalized terminology, without reasoning to generate knowledge not explicit
in the input. Such use can be very helpful, but it does not exploit the
true power of logic-based ontologies. And Cyc has made its development
environment and content available, but I do not categorize that as an
"application", rather as a basis for developing an application.
I believe that we have had sufficient development of ontology
development environments to begin to focus more on actual applications.
Further development of environments and reasoning methods is doubtless
needed, but I think that we now can and should develop more agreement on
content, and for that there is no substitute for real applications to test
the underlying ontological theories. I would suggest that any ontology that
gets public funding for content development must have an on-line demo that
can accept test input from the public and show how the ontology helps to
achieve some useful function. And such demos should be available for any
functional component of the application being developed.
I am still a bit puzzled by the fact that after years of work and over
200 million dollars spent on CALO, there is little available publicly other
than some publications and parts of the ontology they use. I have been told
that more results will be put on-line this spring, but it is not clear
whether any of that will provide us with an ability to determine whether the
money was spent wisely - whether the public got anything usable from the
project. With that level of spending, I think we should have a lot more to
show by this point. The fact that we don't suggests to me that the model
for funding of research in ontology and AI functions that use any form of
knowledge representation is inappropriate for the problem. At this point,
public availability of results and functions at every point in the project
should be mandatory as a condition for public funding (of unclassified
work).
That is one of the basic points I have been trying to make (perhaps
inarticulately) in my suggestion about a new Foundation Ontology project.
Developing any new ontology with content in it only makes sense to me at
this point if it is tied directly to parallel and integrated development of
some application that uses the content. For the Foundation Ontology the
application is not just *one* domain application, but the *integration* of
domain applications - which is why the project has to be of some significant
size. To show that an FO is usable for interoperability, at least two or
three domain applications that use the FO and whose integrated functioning
can be enhanced by the FO need to be developed as demonstration
applications. And the components should be available to the public
individually and in combination as soon as they actually do anything that
can demonstrate their functioning. I also think that a good
natural-language interface will be important to help demonstrate the power
of the ontology - but it may not be essential. (01)
Pat (02)
Patrick Cassidy
MICRA, Inc.
908-561-3416
cell: 908-565-4053
cassidy@xxxxxxxxx (03)
> -----Original Message-----
> From: ontolog-forum-bounces@xxxxxxxxxxxxxxxx [mailto:ontolog-forum-
> bounces@xxxxxxxxxxxxxxxx] On Behalf Of Ed Barkmeyer
> Sent: Monday, April 13, 2009 11:54 AM
> To: John F. Sowa
> Cc: [ontolog-forum]
> Subject: Re: [ontolog-forum] ISO merged ontology effort "MCO"
>
>
> John F. Sowa wrote:
>
> > EB> ... the purpose of an ISO "study period" is to determine whether
> > > there is a specification or parts of a specification that has
> > > sufficient consensus for standardization and meets some perceived
> > > communal need (either in industry, or in the making of other
> > > standards).
> >
> > That is a laudable aim. But those terms 'specification' and 'parts
> > of a specification' are very unclear by themselves.
>
> Not for an ISO study group. A standard is a normative specification
> for
> common practice. The function is to get consistency of industrial
> practice, interoperability of systems and practices, and agreement on
> things that have commercial value, legal significance, etc.
>
> The purpose of a study period is to decide whether there is some
> available specification/document or documents, whose content is ripe
> for
> standardization, that is, on which there is sufficient agreement in
> common practice. The same idea applies to measuring tire treads,
> packing tomatoes, and creating basis ontologies for software
> interactions.
>
> > When it comes
> > to ontologies, people tend to focus on the names of the categories
> > and their placement in some hierarchy.
>
> So that is what they standardize. When it comes to tire treads, people
> tend to focus on remaining lifetime under various road conditions; so
> they standardize measurements that are generally considered to be
> indicative of that.
>
> > But pioneers in the field,
> > such as Aristotle, Leibniz, and Kant, emphasized the *methodology*
> > for deriving the categories.
>
> And I don't doubt that Goodyear and Dunlop and Michelin created the
> methodologies for estimating tire life.
>
> All the rest of this is about HOW practice in ontology development is
> established, and I have no argument with it. But WHETHER there is now
> common practice, or a common result of different practices that can be
> used as a basis for further common or diverse practices, is the
> question
> the study group must answer.
>
> > EB> The idea that some amalgam of BFO, DOLCE and SUO would have
> > > such consensus would be supported by findings from the study
> > > period that some useful common set of concepts is essentially
> > > identical across these ontologies, differing only in terminology.
> >
> > Two points in that comment seem arbitrary: the phrase 'some amalgam'
> > and the choice of BFO, DOLCE, and SUO. As I said before, Cyc is the
> > largest formal ontology on planet earth, and it has undergone almost
> > a quarter century of continuous development with contributions from
> > many highly respected logicians, linguists, philosophers, and experts
> > in artificial intelligence.
>
> But Cyc is not available to those of us without appropriate clearances
> and a "need to know". OpenCyc is what is available in that line. I
> don't know why the U.S. recommendation didn't include OpenCyc. A fair
> guess is that the U.S. contributing body didn't contain an apostle for
> Cyc. And it is certainly not too late to add it to the work list of
> the
> study period, by having someone who knows and cares about it
> participate
> in the study group.
>
> > The word 'amalgam' makes me think of a dentist mixing mercury with
> > powdered silver.
>
> In the manufacturing world, amalgamation is a common means of obtaining
> materials with the desired properties by combining materials that have
> the desired properties severally, or alter the properties of one
> another
> in the desired ways. Unfortunately, amalgamation also tends to beget
> undesirable side effects as well. So I used the word here "with malice
> aforethought". ;-)
>
> I don't take issue with the gist of John's email: how to do ontology
> development. But the ISO issue is not (yet) how to do ontology
> development. It is whether we now, in 2009, have sufficient agreement
> of
> the independently developed foundation ontologies to make one
> consistent
> one that everyone can use a basis for making useful industrial
> ontologies.
>
> That is, at this time, what MCO is about, and that was the topic in the
> subject line. (Not that the topic in the subject line necessarily
> reflects the nature of the discussion on this exploder after the first
> day, but ...)
>
> Now, if John can proffer approaches for judging the quality,
> extensibility, utility, etc., of a "fundamental ontology" (as Chris
> Welty tried to do a few years back), THAT would be (IMO) of direct
> value
> to the study group.
>
> -Ed
>
> --
> Edward J. Barkmeyer Email: edbark@xxxxxxxx
> National Institute of Standards & Technology
> Manufacturing Systems Integration Division
> 100 Bureau Drive, Stop 8263 Tel: +1 301-975-3528
> Gaithersburg, MD 20899-8263 FAX: +1 301-975-4694
>
> "The opinions expressed above do not reflect consensus of NIST,
> and have not been reviewed by any Government authority."
>
> _________________________________________________________________
> 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
> (04)
_________________________________________________________________
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 (05)
|