From: Mike McCarty (jmccarty_at_ssd.usa.alcatel.com)
Date: 2001-01-25 01:34:56
On Thu, 25 Jan 2001, MagerValp wrote:
> >>>>> "MM" == Mike McCarty <jmccarty_at_ssd.usa.alcatel.com> writes:
>
> MM> I know of no good machine translators. They don't exist.
>
> Well, there are some that are good, but they are limited to one
> domain, f ex the Canadian French<->English weather report translator.
There are indeed reasonable translators which limit themselves to
artificial languages. I thought we were talking about natural
languages. I am aware of the canadian weather report translator. But it
does not translate natural language, it translates artificial languages
designed to be compatible with English and French in a narrow field.
> MM> Early in the "machine translation project", people thought that in
> MM> order to translate, not much semantic processing was required. As
> MM> a result, machine translation, like controlled economical fusion,
> MM> has always been 20 years away. It is now understood that
> MM> reasonable translations cannot be achieved without doing
> MM> significant semantic analysis.
>
> Well, they found that out in the 70s.
Yes.
> MM> So much so that no one currently has a clue how to make a computer
> MM> do it.
>
> But people have actively working on semantic and pragmatic analysis
> for decades now. Seems like something should have appeared by now.
Seems so to you. It is a very knotty problem.
> Though last time I checked, brute force translators (such as Systran)
> were the only ones that worked even remotely like they should.
Yep.
> MM> "Fruit flies like a banana."
>
> MM> Is "flies" a noun, or a verb?
>
> Context my friend, context :)
Which requires *semantic* processing, at levels which are unbelievably
deep.
> /Pelle, ex-computational linguist
What is an ex-computation?
:-)
Mike
----
char *p="char *p=%c%s%c;main(){printf(p,34,p,34);}";main(){printf(p,34,p,34);}
This message made from 100% recycled bits.
I can explain it for you, but I can't understand it for you.
I don't speak for Alcatel <- They make me say that.
----------------------------------------------------------------------
To unsubscribe from the list send mail to majordomo_at_musoftware.de with
the string "unsubscribe cc65" in the body(!) of the mail.
This archive was generated by hypermail 2.1.3 : 2001-12-14 22:05:38 CET