[topicmapmail] Re: mulitilingual thesaurus and ontologies

[Mason, James David (MXM)]

OK, Bernard, I'll bite. 

Compared to John Sowa and the things he discusses in his "Knowledge
Representation" (http://www.jfsowa.com/krbook/index.htm), I'm an absolutly
ignorant neophyte in formal systems. I'm  I'll even admit to having dropped
the one college course I started in symbolic logic (some 37 years ago). The
most I can claim is that I've been a comparative linguist, I've masqueraded
as a programmer, and I've watched the development of Topic Maps for some
years.

However, in the work on the Ferret analytical engine that I've spoken about
at a couple of conferences, I've attempted to build a kind of logical
system. (For those who haven't seen the material, the version of the paper
that's about to be published in "Markup Languages" is at
http://www.y12.doe.gov/~mxm/open/Papers/Ferret2001ML.pdf; the presentation
slides from XML Europe are at
http://www.y12.doe.gov/~mxm/open/Papers/Ferret2001.ppt.) What I've done with
Ferret is no doubt quite naive, but I'm willing to toss it into the
discussion in hopes of getting ideas flowing (and perhaps also getting some
constructive criticism of the work).

Jim Mason

James David Mason, Ph.D.

Y-12 National Security Complex
Bldg. 9113, M.S. 8208
P.O. Box 2009
Oak Ridge, Tennessee 37831-8208

+1 865 574 6973

Chairman, ISO/IEC JTC1/SC34
http://www.y12.doe.gov/sgml/sc34/ornlsc34oldhome.htm


> -----Original Message-----
> From:	Bernard Vatant [SMTP:bernard.vatant@mondeca.com]
> Sent:	Tuesday, February 05, 2002 4:51 AM
> To:	John F. Sowa
> Cc:	TopicMapMail
> Subject:	[topicmapmail] Re: mulitilingual thesaurus and ontologies
> 
> [John Sowa]
> 
> > It is true that any particular version of logic must be expressed
> > in some language, such as predicate calculus, CGs, KIF, Aristotelian
> > syllogisms, or even topic maps -- all of them are different notations
> > for some subset of the fundamental principles of logic.
> >
> > But logic itself is a standard that is language independent.  It is
> > the prerequisite for any language that purports to communicate
> > anything about the world or about any viewpoint about the world or
> > about any language of any kind.
> >
> > That is why I say that if you want to use topic maps to communicate
> > different viewpoints about the world or about signs, your first
> > obligation is to determine their logical foundation.
> >
> > Bottom line:  Either topic maps are a notation for some subset
> > of logic, or they are a meaningless notation.
> 
> John
> 
> Without abusing of your time, I think this core debate has to be pushed
> into TM community,
> and I cc this message to the list.
> I had last year basically the same kind of answer in a similar debate with
> Pat Hayes ...
> I think the best way to convince TM people of your point is indeed someone
> really starting
> working on some TM logical model.
> The problem is that we have so far in TM community, nobody that I know
> with enough
> background in formal logic to be able to set the kind of model you seem to
> be asking for
> before even starting to discuss about it. If I am wrong, please volunteers
> move one step
> forward. There is indeed an ongoing work on a "graph model". See TMPM4 at
> www.topicmaps.net. But I guess you won't call that a logical foundation,
> will you?
> 
> And the other way round, academic people who have a strong background in
> logic have so far
> not shown any real interest in topic maps, AFAIK I heard no one saying:
> "Hey folks, I've
> put together some formal model for TM. Does it fit your needs?"
> 
> So I feel like going to the garage because I have some trouble with
> fine-tuning my car
> engine, and hear the chief mechanic say: "Sorry. I don't understand what
> your problem is.
> Please review your engine concept and come back after that". I would
> answer : "Hey, but
> *you* are the engine expert! Can't you help me about it?" And he would
> answer : "No. Your
> engine concept is ill-defined. Set a good model, without which I can't
> even speak about
> your problem".
> 
> See what I mean?
> 
> It's true that there are so many attempts in the "standards" community, in
> so many
> languages and formats, that certainly reinvent many existing wheels. But I
> figure the role
> of academics for that matter could be more to help and provide relevant
> tools, than sort
> of despising those efforts, waiting for them to come to the point they are
> worth
> discussing, or die out ...
> 
> Just thoughts ...
> 
> Bernard
> 
> 
> 
> _______________________________________________
> topicmapmail mailing list
> topicmapmail@infoloom.com
> http://www.infoloom.com/mailman/listinfo/topicmapmail