[NTLK] Improving Cursive recognizer?

From: David-Michael Cook (bluefusion_at_mac.com)
Date: Tue Mar 11 2003 - 18:24:19 PST


I have great luck with the printed recognizer on the Newton, but just
recently became aware of the fact that it does not "learn", and
apparently the Cursive recognizer does. Now, the Print recognizer is
certainly good, but it misses some common things, presumably because I
can't "teach" it.

So I'm trying the Cursive recognizer, and for some reason I'm really
not having good luck with it--no matter what settings I use, it isn't
really anything like what I'm writing half the time. What are some tips
for improving its recognition? I want to see its "learning" in action,
and at the moment it's slightly better than it started out as (about
six hours of writing so far over a few days, but it still has around a
70% accuracy or less, whereas the print recognizer gets a good
85-90%...)

So does anyone know the secret to getting Newton to recognize better?
How to write, etc.? Thanks!

-- 
This is the NewtonTalk list - http://www.newtontalk.net/ for all inquiries
List FAQ/Etiquette/Terms: http://www.newtontalk.net/faq.html
Official Newton FAQ: http://www.chuma.org/newton/faq/


This archive was generated by hypermail 2.1.5 : Wed Mar 12 2003 - 11:30:00 PST