Speech Recognition: An Accommodation Planning Perspective by Alan Cantor may
give some clues
http://www.colorado.edu/sacs/ATconference/Cantor%20-%20Speech%20recognition,
%20handout.htm
I was also going to mention the Ace Centre Book and agree with what has
already been said on the subject. .
Sadly I have never come across a set of forms that would help me through the
assessment process but maybe this is something we should look into as I am
sure it would help others! I have tended to carry out all my assessments
based on the person's personality and willingness to keep trying to work
with the software despite set backs when they have a cold, feel tired and
commands are turning into large paragraphs of rubbish. I am now e-mailing a
colleague in the States as we were just talking about the lack of research
into this aspect of speech recognition work.
As a speech and language therapist I always feel that we do not analyse the
person's language output well enough - knowledge that someone has a
relatively wide range of vocabulary with good sentence construction plus the
ability to dictate at ease into a recorder can help when it comes to the
blank screen. It is also worth seeing how much tension is involved as you
do not want to end up with RSI of the vocal cords because someone is trying
so hard to make the machine cope with their hesitations, pauses and
inability to immediately correct with ease. The training tends to work
because it is read as do the early attempts when you may use some written
notes to get people going but keeping the flow up when composing ad hoc is
very different. Test whether looking at the screen causes distraction and
loss of flow of ideas as it may be better to encourage the user to look away
as they are thinking about what to say.
Best wishes
EA
Mrs E.A. Draffan
Assistive Technologist
Mobile Tel: 07976 289103
|