Hello all!
I've received several responses to my question. I thank you all. Some
of you suggested me to give more information about what I want to do,
what language I want to work with, etc. Well, so that's what I'll do.
I'll talk to you about UTL, my project for multilingualization, in
which there are already several people interested in.
UTL (Universal Translation Language) is the language I want to make a
multilingual translation software for. UTL is not a natural language,
it's an artificial one. It's similar to Esperanto, but UTL has the
difference of being specially suitable for to be machine translated
without mistakes. It's an unambiguous, logical language.
That means it can be a powerful tool for multilingualization of
monolingual texts. A human translator (specialist in UTL) would
translate the original text (written in English, French, etc) into the
UTL language. This UTL version would be given to the translation
program to be converted to several natural languages, with results
that are expectable to be much better than if we made the machine
translation directly from the original text.
So, with this process, we have translated one text to several (15,
20...) other languages at the cost of just one human translation.
This could become a very popular tool in e.g. Internet, where so many
people (specially commercial companies, etc) would like to have their
web pages translated to 10 or 15 languages but cannot afford the costs
of 10 or 15 professional translators.
Some common questions people make are:
1. Is UTL a complete language? Can it express anything that can be
expressed with e.g. English?
The answer is yes. UTL is heavily based on Esperanto, a complete
auxiliary language which have been used by several millions of
speakers around the world for more than a hundred years. Thousands of
books have been translated to it from many different natural
languages. But it's better that you document yourself about this in
the many Esperanto-related websites in the web.
2. Who is going to learn UTL? It's not easy to learn a whole new
language...
Well, the answer to who is going to learn UTL is very simple: anyone
who wants to make money. A UTL translator will be paid as any other
translator for his work. And it's probable that he earns even more,
specially while there be not many UTL specialists available.
Anyway, UTL is more or less as easy to learn as Esperanto, which turns
out to be a few months of study. But the very more probable thing is
that most UTL learners come from the Esperanto speaker community. For
them, learning UTL will be a matter of very few weeks, perhaps days.
3. What makes UTL so suitable for being machine translated?
That's not easy to explain in a few lines, and less to an audience
probably not familiarized with artificial human languages. But we
could say UTL is like a machine translation interlingua but with the
format of a human speakable language.
Well, that's all about it, more or less. I will appreciate any
comments or suggestions.
Greets,
Marcos
PS: For if you were going to ask, yes, I'm acquainted of the DLT
project, which had some things in common with this one. However, as
you can see, both the approach and the goals are different in UTL. I
think this is the first work ever in this direction.
--
For MT-List info, see http://www.eamt.org/mt-list.html
