Quote
Originally posted by Slayer:
(As a Ph.D. student, you absolutely positively HAVE to do something you find interesting).
I would think so, since you'll be working on it for a long time...
Actually, your instructor's idea sounds neat to me, but then, the deaf people can already type just fine, so it's not like the technology would allow them to really do something they couldn't do before. (Albeit possibly slower if they aren't touch typists.)
Can computers already simulate signing on their own? Such a text-to-sign technology would be interesting too, but i don't know that it would be considered challenging enough for a Ph.D. (Film a few people signing with a digital camera, make a dictionary, tah-ta-dah problem solved...)
'Course, sign language is pretty complex, at least ASL, from what I've heard. That might actually be pretty hard...
Ah well. I hope you find something of great interest to you. It can certainly be hard to find something that stays interesting.
I say, go with machine learning. When I read it someday, I want to be able to have a chance of understanding it, and I'm not much of a mathematician
Maybe make a program that crawls the web like goggle, but makes up a database of knowledge rather than web sites. You might ask it:
Can penguins fly?
And it would match that with occurrences of 'penguin' and 'fly' in it's database (as well as variants on fly such as flightless) and then return:
The penguin is a flightless bird
Or something like that.
Ah well, i'm sure you could think of a better idea anyway... besides, the computer couldn't tell between frequently-posted fiction and factual information... or could it?
Anyway, Good fortune to you in this project.
------------------
"Programming is an art form that fights back."
-
Unknown
"I do not fear your Powerful Racoon, for I have a Short Poker of Shooting!" -
Random Story Creator IV