September 02, 2008

reCAPTCHA helps libraries and archives take over the world

The most exciting news I read all last week had to do with the use of reCAPTCHA internet security software in transcribing tricky, hard to read words in on-line copies of newspapers and old books. CAPTCHA is responsible for the squiggly security codes humans are asked to transcribe on websites, authenticating themselves as humans and not machines. Now this idea is being put to a new use, helping out where OCR fails. Those words OCR can’t read, because they’re smudged, weird, cramped, strangely-fonted, badly-spelled, whatever, can be forwarded to reCAPTCHA and it will be transcribed. Here’s their website: http://recaptcha.net/learnmore.html

No comments: