I am gonna step way out of my mathematical depth here
I mean no disrespect, but this is the last accurate statement you made.
You are probably correct, I am certainly out of my field and hope
that noone would embark on a hopeless project based on my few words.
However, the school district (and some teachers) tell me that
Rosetta Stone does this quite well (phrasal pronunciation scoring).
I've heard there are other packages that do as well
This tells me the math has been done.
But you are correct that he would not want to attempt the math himself,
I just think it can probably be found somewhere and MatLab users might
be a good starting point.
Personally, I have done from scratch lots of things which were done
before and sometimes better.
Anyway, it is a truly fascinating subject.
The true test of whether you are learning a language is whether
native speakers can understand you well, not some computer graded
pronunciation test. The pattern fitting capabilities of the auditory
regions of the neo-cortex and limbic regions are so awesome we may be
decades away from automating them. Emulating hundreds of thousands of
synaptic connections and meta-language itself is gonna take a lot of
work.
I will defend one statement though. I have yet to see anything which
Python would not make a good wrapper for. Some of the OpenGL pygame
stuff is very cool.
