[Top] [All Lists]

Re: [ontolog-forum] web-syllogism-and-worldview

To: "[ontolog-forum]" <ontolog-forum@xxxxxxxxxxxxxxxx>
From: "John F. Sowa" <sowa@xxxxxxxxxxx>
Date: Tue, 21 Apr 2009 01:52:17 -0400
Message-id: <49ED5F11.90105@xxxxxxxxxxx>
Dick, Duane, and Pat,    (01)

RHM> From what I've read about childhood development,
 > young children use a very simple language of their own.
 > Words in a sentence may be in no particular order, or
 > they may use a simple "pivot" grammar.    (02)

That's true of the earliest stages.  But following are some
sentences by a child named Laura at age 2 years, 10 months:    (03)

    Here's a seat. It must be mine if it's a little one.
    I went to the aquarium and saw the fish.
    I want this doll because she's big.
    When I was a little girl, I could go "geek geek" like that,
    but now I can go "This is a chair."    (04)

http://pubpages.unh.edu/~jel/JLimber/Genesis_complex_sentences.pdf    (05)

By the way, Laura was a bright child whose language was being
studied by a bunch of psycholinguists.  So she might be somewhat
more advanced than most children, but nearly all of them reach
a similar stage between the age of 3 and 4.    (06)

DN> Every coffee table sold at Iowa has four legs. Every dog is
 > born with four legs. Unless you can disambiguate that legs has
 > a different meaning in those two np contexts, the logical
 > conclusion would be inaccurate.    (07)

That is true.  The lexical resources are adequate to parse the
sentences, but you can't do detailed reasoning unless you
determine the specific domain and use the detailed axioms
(i.e., microtheory) for that domain.    (08)

PC> I agree with most of what John says, except for the 'fractal'
 > part.  Ontology is certainly *capable* of being extended to
 > indefinite levels of detail,    (09)

That is what fractal means -- the open-ended possibility of
an overwhelming amount of detail at every level.    (010)

PC> but practical applications do not require indefinite levels
 > of detail.  One needs to describe all the detail that is important
 > for the application at hand, and that is enough.  If more detail
 > is needed, more detail should be added.    (011)

That is indeed what I would recommend.  But I would recommend it
"just in time."  Don't develop any more detail than necessary.
But when you do develop it, save it for possible reuse later.    (012)

(And by the way, the hierarchy of theories is a good place to
save it and to look for theories that other people have saved.)    (013)

PC> ... the 'synsets' derivable from it would *look* like WordNet
 > synsets.  It would have support for much greater levels of
 > semantic detail and semantic precision, and would include
 > important rules and functional representations that WordNet
 > cannot represent.    (014)

But I would split the two.  The synsets belong to the lexical
resources.  The precise details belong to the domain ontologies,
which correspond to Cyc's microtheories.  We do that in the
applications we build at VivoMind.  For examples, see    (015)

    http://www.jfsowa.com/talks/pursue.pdf    (016)

The last slide of that talk has pointers to other references.    (017)

PC> What I am trying to do now is to test the use of an ontology,
 > with the ontology elements serving as 'senses' and specifying
 > which linguistic labels (English words or phrases) are used
 > to refer to those concepts in ordinary language.    (018)

That's fine.  But each word sense in English corresponds to
an open-ended number of microsenses for each and every domain.
It's impossible to associate axioms with the lexical terms
because their precise definitions in different domains would
be inconsistent with one another.    (019)

PC> I have to confine my own efforts to the basic vocabulary
 > of 2000-5000 words.    (020)

Why restrict your efforts to such a tiny vocabulary?  At
VivoMind, we use multiple off-the-shelf lexical resources,
such as WordNet or Roget's Thesaurus or anything else that
anyone might develop.  Those resources are primarily used
to resolve ambiguities in the English text.  Since each
word sense is so highly ambiguous, there's no need to
worry about the precise definitions.  The precise level
of reasoning is done in the domain ontologies (AKA the
microtheories in Cyc's terminologies).    (021)

For somewhat more detail about how we handle that, see    (022)

    http://www.jfsowa.com/pubs/pursuing.pdf    (023)

PC> The existing WordNet synsets are excellent references
 > and resources, but one cannot take them as coherent
 > ontological concepts, though that is the way that some
 > people attempt to use them.    (024)

I certainly agree.  The lexical resources (such as WordNet)
are useful for resolving most of the ambiguities in the
English text.  But the precise semantics and the detailed
reasoning is done with the domain ontologies.    (025)

John    (026)

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    (027)

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