This is an interesting application, Tony. Your idea is more
likely to succeed than some of the others, because you're not trying to do
too much at once. Several projects try to go from speech to sign and back
right away!
What you're really talking about is a machine translation system,
like a portable version of the Babelfish (https://babel.altavista.com).
You should know right up front that it's a very difficult task. I did a
prototype with English to American Sign Language (one way, non-real-time,
very limited domain), that you can read about on
<https://www.unm.edu/~grvsmth/portfolio/>.
I'd recommend starting with a general introduction to machine
translation. And since you don't know a sign language yet, you might want
to try your project with a language that you do know -- maybe Swahili to
English, and then use the same technology for Irish Sign Language (is that
what's spoken in Northern Ireland, or is it British Sign Language?
Michael?), or whichever sign language you choose.
There are some other SignWriting-related projects that may not be
quite as flashy, but still need someone to spend a lot of time on
them. Top of the list in my mind is that the way of typing signs that was
invented for SignWriter is not very intuitive or comfortable. A thorough
human factors analysis would do wonders by improving the speed and ease of
typing in SignWriting.
--
-Angus B. Grieve-Smith
Linguistics Department
University of New Mexico
|
SignWriting List Forum