Obrst, Leo J. | 16 Dec 18:19 2007
Picon

CFP: Interdisciplinary Ontology Conference, Tokyo, February 26-27, 2008


Apologies for cross-postings and duplications. Please forward to
colleagues. 

*******************************************************************
    Interdisciplinary Ontology Conference (InterOntology08 Tokyo)
*******************************************************************
We are pleased to announce an international conference to be held in
Tokyo on February 26-27, 2008. The conference will be co-organized and
co-sponsored by

  - JCOR: the Japanese Center for Ontological Research
  - NCOR: the (US) National Center for Ontological Research
  - ECOR: the European Center for Ontological Research

On the afternoon of Tuesday February 26 a special session will be held
on Biomedical Ontology, jointly organized by JCOR and the (US) National
Center for Biomedical Ontology.

This conference will serve to launch JCOR (the Japanese Centre for
Ontological Research), which has recently been funded by the Japanese
Government's Ministry of Education and Science (MEXT) under the
framework of the Open Research Centre on Logic and Formal Ontology.

The conference will be held at Keio University, Tokyo.

The invited speakers include;
  - Alan Ruttenberg (USA)
  - Werner Ceusters (USA)
  - Nicola Garino (Italy)
(Continue reading)

Jon Awbrey | 21 Dec 16:08 2007
Picon
Picon

Re: AND/OR Search Spaces in Graphical Models

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o

Rich, CG List,

There are inherent conceptual and structural inefficiencies in the use
of AND/OR graphs and the associated search spaces, like those implicit
in Peirce's alpha structures.  Those inefficiencies can be alleviated
to a considerable degree by moving on to the generalization of trees
that graph theorists call "cacti".

When I began my efforts to implement Peirce's logical graphs on
the computer back in the mid 70's -- does anyone remember SNOBOL? --
I quickly ran into the computational side of these inefficiencies.
By the mid 80's I had learned enough about cacti to have programmed
a few prototype logical utilities that implemented generalizations
of Peirce's basic ideas.

Once you take the step into the cactus patch, many collateral
advantages of using cactoid structures quickly become apparent.

Some of this work is documented -- or in the process of being documented --
on assorted pages of a wiki-based directory, for instance, these pages:

http://www.mywikibiz.com/Directory:Jon_Awbrey/Projects/Cactus_Language
http://www.mywikibiz.com/Directory:Jon_Awbrey/Papers/Futures_Of_Logical_Graphs
http://www.mywikibiz.com/Directory:Jon_Awbrey/Projects/Theme_One_Program

Incidentally, this site is open to anyone and implements a very nice variation
on the usual wiki protocol, one that allows both collaborative work (in the main
space on the discussion pages) while preserving the proprieties of "owned" work
(Continue reading)

Rich Cooper | 27 Dec 18:41 2007

No poverty of the stimulus

Not everyone seems to believe that the "poverty of the stimulus" is a valid
argument.  

Here is a mathematically supported paper that provides a deep treatment of
learning language
"Ideal Learning of Natural Language: Positive Results from Learning about
Positive Evidence"
by Nick Chater at Univ Coll London:
http://eprints.pascal-network.org/archive/00002798/01/jmp06.pdf

His claim is that an ideal learner, using Kolmogorov complexity methods, can
provide both the positive and the negative learning needed for an ideal
learner, even though only given positive evidence.  The absence of evidence
for certain constructions is treated as evidence that those constructions
are ungrammatical.  

Comments appreciated.  

Sincerely,
Rich Cooper
http://www.EnglishLogicKernel.com

_______________________________________________
Corpora mailing list
Corpora <at> uib.no
http://mailman.uib.no/listinfo/corpora

P Resnik | 27 Dec 22:58 2007
Picon

Re: No poverty of the stimulus

Rich, thank you for posting this pointer!  I have not yet had a chance to read the paper, but on a quick scan, I notice that they do not reference

Shyam Kapur , "Computational Learning of Languages",  Cornell dissertation, 1991.  http://portal.acm.org/citation.cfm?id=866568 .  Postscript and PDF available at http://ecommons.library.cornell.edu/handle/1813/7074

I suspect that anyone interested in the Chater and Vitanyi paper will also find this of interest, particularly the results of Chapter 5, where Kapur writes, "... in accordance to a suggestion due to Gold (1967), maybe we can learn more families if we insist on convergence on most (instead of all) texts [i.e. sequences of sentences presented as positive examples to the learner -PSR]".  His work shows that with this convergence criterion, it is possible to obtain "a uniform learning algorithm that works for every family of languages" [my emphasis], subject to stochastic assumptions about the input that, if I understood/recall correctly, avoid problems with the sorts of pathological texts the Gold's proof relied on.

Happy holidays,

  Philip
`


On Dec 27, 2007 12:41 PM, Rich Cooper < Rich <at> englishlogickernel.com> wrote:
Not everyone seems to believe that the "poverty of the stimulus" is a valid
argument.

Here is a mathematically supported paper that provides a deep treatment of
learning language
"Ideal Learning of Natural Language: Positive Results from Learning about
Positive Evidence"
by Nick Chater at Univ Coll London:
http://eprints.pascal-network.org/archive/00002798/01/jmp06.pdf

His claim is that an ideal learner, using Kolmogorov complexity methods, can
provide both the positive and the negative learning needed for an ideal
learner, even though only given positive evidence.  The absence of evidence
for certain constructions is treated as evidence that those constructions
are ungrammatical.

Comments appreciated.

Sincerely,
Rich Cooper
http://www.EnglishLogicKernel.com





_______________________________________________
Corpora mailing list
Corpora <at> uib.no
http://mailman.uib.no/listinfo/corpora


_______________________________________________
Corpora mailing list
Corpora <at> uib.no
http://mailman.uib.no/listinfo/corpora

Gmane