I am doing research on medication errors. In the US (and elsewhere) the
majority of prescriptions are still handwritten. Physicians' handwriting is
notoriously illegible. Does anyone know of a corpus of handwritten text. I am
especially interested in legibility and error data, e.g., what handwritten
letters are most likely to be confused with one another. What words or letter
patterns are most likely to be misread? People working in OCR may have insights
into this issue as well. I'd appreciate any pointers.
Thanks in advance.
Bruce Lambert
University of Illinois at Chicago