Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: Re: SUO: Vote 2001-02: IFF Foundation Ontology




Hi,

At 07:49 19/08/01 -0400, yangyung wrote
I think I understand the example.

There are 2 games here:
- the physical moves in the tower of hanoi puzzle
- the mental moves in the game of choosing 'what to do next'

It's right

By parsing verbal statements you hope to make a judgement
about what mental moves the player has made or will make.

yes


[1]The physical moves can be coded as pairs of states eg
(4321 0  0  , 432 1 0)

Yes it can be

[2]For each pair in [1] you know can determine what strategies
are theoretically feasible. Other possibilities would be found in
practice, such as making illegal moves . THese possibilities
can be further classified as either moving-toward or
moving-away from a solution.

yes, one can established some variables in order to depict the succession 
of moves for a strategy, and their position in relation to the optimal 
strategy. I tried to do it. I did something.


[3]A verbal description about such a small universe of objects
can be reduced to a small number of different possibilities
using standard Prolog NL parsing and reduction techniques,
and these in turn can be matched with the possibilities from
[2] , further reducing the number of solutions.

Here is a step of the work that I can't do myself because I am not a 
computer specialist and I do not know how to use PROLOG
Furthermore it can be true at a particular level of analysis, I think of 
the level of propositions (or relations)

But there are other levels of language analysis, for exemple a pragmatic 
one (for me it is enunciative level). What is relevant at pragmatic 
level  is the way in which things are said for every sentences. So 
differences have to be introduced between the proposition level and the 
utterance at every step of the strategy.
For example functional markers such as connectives, thematic organisation, 
modal markers lent me to introduce these differences.
So from the organization of the linguistic markers  one can explicit the on 
line organisation of mental moves and of mental representation and mainly 
the way in which on-line reorganizations occured.
My assumption is that some elementary processes are invariant


The sentences would be distinguished into explanations of
what had happened, or plans of what to do next.

Yes
Moreover at pragmatic level differenciations have to be done.
So one can have a pragmatic approach of problem solving area. It implies 
that new concepts have to be elaborated in this area to give an account of 
what could be a general problem solving strategy from the point of view of 
a pragmatic approach of verbal reports.

[4] The precise format would depend on your research parameters.
My understanding above yields:

- pair of physical move states from [1]
- from [2], for each pair there can be list of possible strategies
   of how to compose Towers-of-hanoi moves into larger sequences.
   The recursive nature of the problem means these can be
   composed into higher and higher strategies.


Disks are also composed in larger and more or less structured objects (I 
can find them from the linguistic organization)


But novice can't solve the problem recursively without to learn a lot of 
things. It is this learning process I try to study.
Even with experts there are not higher and higher strategy because at a 
specific level there are very parsimonious strategies that they prefer.
For example they can construct automatizations to move (in one single unit) 
two disks from a peg to another one (2-disks moves), and idem for three 
disks (3-disks moves). But they prefer to decompose the 3-disks moves into 
two structured 2-disks moves (a first 2-disks moves for spreading the 
disks, another one for the reconstruction of the pyramid after disk3 has 
moved). In fact they don't like very much the recursive strategy.
One can have these kind of results with the pragmatic approach of language.

- a verbal statement
- a parse tree of the the verbal statement in Prolog
- a reduction of the parse tree , giving possible matches
   of strategies thatthe player has done or will do

yy

At a first level yes.
But I don't know how is the form the verbal statement in PROLOG
The different approaches of a statement have to be matched with the 
differentiations which have to be introduced at the problem solving level.
But of course, it is easier at the problem learning level, especialy in 
case of the well known problem of tower of Hanoï
All things have to be O.K. at level of problem solving even if we don't 
know every thing at language production level.

Josiane
_________________________________________________

Josiane Caron-Pargue
Laboratoire Langage et Cognition LaCo, UMR 6096
Maison des Sciences de l'Homme et de la Société, MSHS
99 Avenue du Recteur Pineau, F-86022 Poitiers cedex
tel bur 05 49 45 46 23, fax 05 49 45 46 16
tel secr. 05 49 45 46 10, 05 49 45 46 13
http://www.mshs.univ-poitiers.fr/laco/
josiane.caron@mshs.univ-poitiers.fr
_________________________________________________