An adaptive handwriting recognition system1

Abstract

This paper describes a recognition system for on-line cursive handwriting that requires very little initial training and that rapidly learns, and adapts to, the handwriting style of a user. Key features are a shape analysis algorithm that determines shapes in handwritten words, a linear segmentation algorithm that matches characters identified in handwritten words to characters of candidate words, and a learning algorithm that adapts to the user writing style. Using a lexicon with 10K words, the system achieved an average recognition rate of 81.3% for top choice and 91.7% for the top three choices.