[sci.lang] CAI setups for FLL

ruslan@ecsvax.UUCP (Robin C. LaPasha) (03/06/88)

I've recently been musing over certain possibilities for CAI software
based on limited parsing techniques and wildcarding.

Is anyone out there working on software for foreign language learning
that works on the basis of variable student input; i.e. the student
answers a question or translates a sentence, going beyond
fill-in-the-blank with a particular word?

I've read of peripheral stuff; some ESL work in _Computers in the Schools_,
parsing for errors in texts in _Computers and the Humanities, and
foreign language text-adventure games in _CALICO Journal_.

But I'm interested not so much in a "how it works in our classrooms"
description of the particular application (I have my own in mind)
as much as how one sets up a program to handle the input - vocabulary,
conjugation and (as applicable to the foreign language) declension,
and variety of syntax.

Besides that, what languages do folks choose to write CAI/CALL in
these days?  I remember PILOT (ugh) and I've looked at CALIS, but I'm
thinking more that a ground-up system in, oh, LISP or Prolog or
whatever seems to work would be much better than trying to fuss
with what I consider as rather unsatisfactory authoring systems.

Can anyone point me to some down-to-earth books/articles/whatever?
(I have read a couple of books about AI and natural language
processing, but, well, I think I need to start at some level
that is practical and simple.  I am neither a hot AI programmer
nor a linguist; I'm a TA of a foreign language with some ideas
on how I'd like to see a CAI system run.)

Would anyone like to discuss such stuff on the net or by email?
If so, let me know by email (and advise me which newsgroup you
intend to discuss the matter in, as I don't regularly follow
all of these groups.)

Thanks in advance,

Robin LaPasha     ruslan@ecsvax.UUCP