NetNews Usenet Archive 1992 #26/NN_1992_26.iso/spool/comp/ai/edu/527

Text File | 1992-11-05 | 9.8 KB | 216 lines

Newsgroups: comp.ai.edu
Path: sparky!uunet!spool.mu.edu!yale.edu!ira.uka.de!chx400!news.unige.ch!divsun.unige.ch!schneide
From: [e-mail] (Schneider Daniel)
Subject: Re: Purpose of this group? - YEAH let's have some AI&ED !
Message-ID: <[e-mail]>
Sender: [e-mail]
Organization: University of Geneva, Switzerland
References: <[e-mail]> <[e-mail]>
Date: Thu, 5 Nov 1992 09:31:52 GMT
Lines: 204

In article <[e-mail]>, [e-mail] (Kemi Jona) writes:
> In article <[e-mail]>,
> [e-mail] (Marie Bienkowski) wrote:
> >
> > I started reading this newsgroup because the name looked interesting.
> > However, I have a question.  Is this a newsgroup 1) to educate people about
> > AI, 2) to talk about AI education (as in "What are the best grad schools
> > to study X?") or 3) to talk about using AI to educate (as in intelligent
> > tutoring systems.
>
> I think Marie's question is telling on a deeper level.  Where is the "real"
> traffic on this group (aside from questions about grad schools and job
> postings)?  Is no one doing AI in education?  Or does no one think there
> are any interesting issues to discuss and debate?  Or perhaps everyone too
> busy building systems to spend time reading this group!
>
> Having said that, I'll step up to the plate with a question:
>
> Does anyone know about any real, working educational software that
> explicitly represents and uses general teaching strategies to conduct
> tutorial interactions with a student?  What have been the roadblocks in
> acquiring and applying such strategies?  What sort of representational
> vocabulary was developed. Any info, or pointers to articles will be
> appreciated.
>
> Come on folks, I know we have stuff to talk about here.
>
> --Kemi


WE certainly haven't got real, working educational software. There are lots
of Roadblocks. One of them is that you need at least 4 people working
full time for several years to get somewhere. An other is that you won't
be able to publish something about having spent those years to implement some
software. An other is that you need interdisciplinary work, something
which is almost impossible to implement in our institutions. And finally,
hard&deep AI&Ed is just to difficult. You run into combinatorial explosions
very fast. You can spend a year on figuring out how students implement
a simple pascal procedure which adds up a few numbers and filters some of
them ...and then you got to hack it up. The future may be "light" Intelligent
Interactive Learning Environments, where we do what we can (an even
that is hard).

We got something we can share. It's mostly theory actually. The
tools are not really ready yet, but they allowed us to explore new ideas
and probably will do so in the future.


ETOILE - Experimental Toolbox for Intelligent Learning Environments

                      (Beta Release 0.9)

The code actually is a early beta release.  Most our efforts went into
conceptual work. The area worst hit by our lack of ressources is the
interface written in CLIM. It's a multi-frame thing (lots of windows
and menus) which currently has big display problems on the Sparcs. I
hardly ever managed to get correct incremental redisplay in some place
... probably because too many things happen at the same time in the
system. There is a coach, a tutor and an expert who fight for
ressources. Stuff you may be interested in is an object oriented
inference engine (tighly coupled to CLOS), an integrated hyper text
(implementing a tiny subset of Latex). Then you might also like our
way of playing with multiple frames, command tables on so forth. We
got a challenge, but no solution there....

You need:
  (1) Common Lisp + CLIM 1.1 (Franz Allegro CL or Genera)
      The code should be easibly portable to other lisps who implement
      processes since it runs on both ACL and Genera. The full system
      has not been used too often on our old Symbolics 20 which is
      basically used to program the interface.

  (2) Megabytes is a plus


This software can be used free of any charge, but we keep the copyright.

Check into our anonymous ftp server:         fpssun19.unige.ch
                                             (129.194.81.43)

                             directory:      /pub/etoile

You can take either take the compressed tar (1.3 MB) or single files
from the ILE directory. (3.8 MB). Both include the same stuff.
Documentation is available in the the ILE/etoile/doc and in the
ILE/memolab/hyp directories. Ftp to Geneva should be fast from
all over the world.


Please ....

 - if by any chance you were able to install and even to run the
   thing, please let me know. We'll try to make things better
   (documentation, code and everything)

 - if you feel an urge to improve the interface, we may be able
   to offer you a RA job for a few month or more in 1993. We may
   get a full year of funding for that if we are lucky.
   Please contact with me, if interessted.
                                                             - cheers Daniel


---------------------------- Abstract from our latest report ------------------


                   NFP23 Program; Project No. 4023-27013
"Les systemes explorateurs intelligents"; Prof. P Mendelsohn and D.K.Schneider.

                     Intelligent Learning Environments

                              August 1992

      B.BORCIC, P. DILLENBOURG, M. HILARIO, P. MENDELSOHN, D. SCHNEIDER.


                            TECFA
            University of Geneva
Faculte de Psychologie et des Sciences de l'Education


Abstract

The design of educational software has evolved during three decades,
reflecting technical advances and changes in theories of
instruction. In this project, we implemented some features of advanced
interactive learning environments (ILEs): the multiplicity of teaching
styles (the same content may be taught in several ways), the
multiplicity of learning sources (experience, coaching, hypertext
browsing), the use of a rich interface allowing direct manipulation
and free exploration, the use of artificial intelligence techniques
for creating complex problem situations and for supporting pedagogical
reasoning.

We developed two pieces of software: ETOILE and MEMOLAB. The first
system is a generic software, an Experimental Toolbox for Interactive
Learning Environments. It provides designers with a set of tools and
resources allowing them to create their own ILE. Globally, ETOILE
provides the pedagogical architecture while the designer provides the
domain expertise and creates the specific parts of the interface.
ETOILE is not an authoring tool that can be mastered by any genuine
author, it is a toolbox to be used by designers with programming
skills in Common Lisp. ETOILE namely includes an Object-Oriented
Production System, a hypertext, various interface management
facilities and pedagogical knowledge bases. Its conceptual
architecture is explained in the User Guide.

MEMOLAB is a learning environment for acquiring basic skills in
experimentation methodology for human sciences. It illustrates the
kind of systems that can be designed with ETOILE.  MEMOLAB includes
several components. The `lab' allows learners to build an experiment.
The simulation produces the results of this experiment. The `data
tools' allow the learner to visualize the results and to compute
basic statistics. During their activities, learners receive the
assistance of a tutor. The system includes five different tutors,
implementing various teaching styles. The learner may also explore the
various hypertextes in which he can find knowledge about designing
experiments and knowledge about human memory.

The design of MEMOLAB reflects our efforts to translate a theory of
development in terms of systems specifications, namely to structure a
learning session into a sequence of micro-worlds, each microworld
being characterized by a specific interface language. The
psychological theory is encoded in the relationship between the
language of successive microworlds. This architecture of a learning
environment is supported by the knowledge structures used in ETOILE.

TECFA is an academic team active in the field of educational
technology. It belongs to the School of Education and Psychology of
the University of Geneva. It is directed by Professor Patrick
Mendelsohn and includes about 10 collaborators. The team is in charge
of the computer courses within the school and provides services within
the computer-based learning community. TECFA main research interests
are the applications of artificial intelligence to education, the
cognitive effects of educational software and the communication issues
with new technologies (distance education, multi-media systems,...).

This research project is funded by the Swiss Research Fund under a
special research program (PNR 23), entitled `Artificial Intelligence
and Robotics'. The contractors are Prof. P. Mendelsohn and D.
Schneider. The research team is multidisciplinary. It includes P.
Mendelsohn (psychology), B. Borcic (computer science), M. Hilario
(computer science), D. Schneider (political and computer sciences), P.
Dillenbourg (educational and computer sciences).

------------

Daniel K.Schneider, TECFA (Educational Technologies and Learning)
Faculte de Psychologie et des Sciences de l'Education, University of Geneva,
9 route de Drize, 1227 Carouge(Switzerland),
Tel.(..41) 22 705 9694, Fax. (..41) 342 89 24

Internet:   [e-mail]  (and various national nets)    | if reply
CSnet/ARPA: [e-mail]   (old style)      | does not
X400:       S=schneide;OU=divsun;O=unige;PRMD=switch;ADMD=arcom;C=ch | work,
uucp:       mcvax!cui!divsun.unige.ch!shneider                       | try one
BITNET:     schneide@cgeuge51                                        | of
DECNET:     UGUN2A::SCHNEIDE (local Swiss)                           | these