[Top] [All Lists]

Re: [ontolog-forum] Data Models v. Ontologies (again)

To: "[ontolog-forum]" <ontolog-forum@xxxxxxxxxxxxxxxx>
From: Ed Barkmeyer <edbark@xxxxxxxx>
Date: Thu, 29 May 2008 12:06:32 -0400
Message-id: <483ED488.2060805@xxxxxxxx>
John F. Sowa wrote:    (01)

> I would say that the work of good lexicographers is far superior to
> the accuracy that has been achieved in the great majority of things
> called "ontologies" that one can find on the WWW.    (02)

Ah, yes.  The work of *good* X is far superior to the work of *average* 
Y.  Where the "average" includes a lot of incompetents with the title.    (03)

The work of *good* knowledge engineers is also far superior to the work 
of a lot of marginally competent automotive engineers.  You just don't 
see the latter work on the Web on in your vehicle, because automotive 
engineering has a testing and review process.    (04)

> EB> The knowledge engineer of 2008 is in the shoes of the structural
>  > engineer of 1500.
> That is a slur on the engineers of 1500.  I suggest you study the
> work of Leonardo da Vinci (1452-1519).  Admittedly, Leonardo was
> a genius, but there were quite a few of his contemporaries who
> weren't bad at all.  They were called "Renaissance men".      (05)

Oh, and were they the average engineers of their time?  I suggest you 
study the work of Mark Musen and the Stanford gang, who have produced 
and collected some very careful and well-thought-out biomedical 
ontologies.  They are contemporary knowledge engineers, who "aren't bad 
at all".  And I could name a dozen others, many of whom are probably on 
this exploder.  So let's try to compare apples with apples, shall we?    (06)

The junk that is on the Web will quickly achieve the oblivion that it 
deserves, along with all the 16th century projects that failed or fell 
down in a few years.  But unlike the 16th century projects, there is a 
lot more visible junk on the Web because it is inexpensive to produce, 
and because it is on the Web and not buried in the archives of Bloedendorf.    (07)

> I would claim that a knowledge engineer today who had a good background
> in logic, linguistics, lexicography, and, I dare say, philosophy, could
> do very good work.     (08)

Well, that probably limits the field to you and Charles Peirce, John. ;-)    (09)

I agree that a background in (formal) logic is required.    (010)

For some applications, the other critical requirement is the 
understanding of how to extract knowledge from published text.  And in 
that area linguistics is indeed very useful, and lexicography may be. 
But some knowledge of the domain itself -- the implicit knowledge of the 
authors and the presumed audience -- is also required.    (011)

For other applications, the critical requirement is the understanding of 
how to extract knowledge from people.  And in that area, a background in 
systems analysis, which includes information modeling and process 
modeling is very useful, and some background in management sciences may 
also be. (There is a kind of "practical linguistics" that emerges from 
that background, but I'm not sure how it relates to the formal 
discipline.) And for the record, people with that kind of background 
also do very good work, but because of its purpose and nature, very 
little of it gets published on the Web.    (012)

And I won't deny that philosophy is generally useful, if only in opening 
the engineer's mind, and avoiding certain common mindsets, like belief 
in universal truth or blind trust in perceptions, which interfere with 
capturing the knowledge/belief of the domain experts.    (013)

But Len is right, even the good work here is not deeply based on 
"theory"; it is based on a knowledge of effective practice.    (014)

> But it's very time consuming, very painstaking work.    (015)

As Tom Edison observed of another engineering trade in its infancy.  But 
I would go so far as to say that "1% inspiration and 99% perspiration" 
is the nature of good engineering.    (016)

> I believe it's possible to speed up that work and to enable people
> with somewhat less stellar qualifications to do decent work.  But that
> would require suitable tools, languages, resources, and analysis aids.    (017)

I fully agree with this, but you left out the most important thing: the 
*discipline*.  A defined, established practice, a set of things you 
always do, and the reason for them, the way the tools and resources 
should be used, the problems and tradeoffs that must be considered, the 
kind of knowledge that must be employed, a process of design, coupled 
with a process of verification.   That is *engineering*, and having and 
using that discipline is what will "enable people with somewhat less 
stellar qualifications to do decent work".    (018)

And you also left out Len's point:  a reference theory, the ontological 
DNA, an understanding of what "captured knowledge" is, and how it can be 
built/evolved from other knowledge, what can be integrated, what can be 
corrected, and what are intrinsically incompatible "microtheories"; the 
paradigms that can be accepted without question and are tests of 
goodness and validity.  We are nowhere near that now, but it is not 
unreasonable to suppose that we might be there in 20 years, or 50, but 
it might also be 300.    (019)

> But that is a topic for a whole 'nuther thread.    (020)

Now there we disagree.  I thought that was the topic of *this* thread!    (021)

-Ed    (022)

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    (023)

"The opinions expressed above do not reflect consensus of NIST,
  and have not been reviewed by any Government authority."    (024)

Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Subscribe/Config: 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 Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx    (025)

<Prev in Thread] Current Thread [Next in Thread>