On 25/09/2004 20:53, Doug Ewell wrote:
Peter Kirk <peterkirk at qaya dot org> wrote:
If we are considering a scenario in which someone takes shorthand
notes at a meeting and transcribes them later, interchange between
computers is likely to be required. If this process is to be
automated, a sensible way to do so would be for the minute-taker to
write shorthand on to a hand-held computer's screen. This data would
then need to be transferred to a desktop or networked machine for
transcription and further editing. The most processing-efficient way
to do so might be to transfer images, but that would be a lot of data
for a record of an entire meeting, so for bandwidth efficiency the
hand-held computer should analyse the shorthand and transfer the
shorthand text in some kind of encoded form.
Is this a realistic scenario? (This is not a rhetorical question; I
really want to know.)
If this sort of workflow actually happens, or would if the encoding were
there, then that might constitute a valid use case for encoding
shorthand. But as far as I know, accurate handwriting analysis of
shorthand or other "normal" penmanship (not Palm Graffiti or a similar
scheme) by commonly available handhelds is not up to this challenge. I
may be wrong.
All I can say is that there is work on progress on automatic recognition
and transcription of shorthand on handheld computers. See for example
http://www.ntu.edu.sg/home/asgleedham/CurrentResearch.htm#Proj6, which
also gives a good justification for implementing such a system, and
http://www.cs.nott.ac.uk/~ceilidh/people.html (last section).
--
Peter Kirk
[EMAIL PROTECTED] (personal)
[EMAIL PROTECTED] (work)
http://www.qaya.org/
