keith said:
> The method is not new. It was used
> successfully for wheather reports already
the "method" might not be new.
but what _is_ different, and undeniably so,
is that google has a _huge_ corpus of text
with which to implement the method now,
possibly the "secret sauce" to make it work.
Just the opposite is the case. Believe me as a computer
linguist. For decades it said that with faster computers
bigger corpora MT would have its break through.
What has happened. Vaporware and results.
It simply does not work. Language can not be
sucessfully model. Languages are regularly formed,
nor well formed.
this asset, and its bearing on the problem,
should not be underestimated. and indeed,
that huge corpus could exert all manner of
effects on a wide variety of knowledge tasks.
the information about the world represented
by _billions_ of web-pages out in cyberspace
could lead to the gleaning of vast knowledge.
(so much so that it could become very scary.)
All AI projects so far have failed and failure has
been admitted. That knowlege can be extracted
from corpora. Language does not constitute
meaning or knowledge. It just transport it.
That is why a good deal in NLP is done in the field
of knowledge representation.
Do you realize that voilets where originally the
color BROWN and not blue !!! (see Goethe).
A translator today would translate Goethe braun(brown)
to blue since it is what people would expect!!!
Keith.
-bowerbird
_______________________________________________
gutvol-d mailing list