reCAPTCHA - Fights Spam / Helps to Digitize Old Books

A CAPTCHA is defined as a, "Completely Automated Public Turing test to tell Computers and Humans Apart."

CAPTCHA's are used to help fight automated abuse of online systems. Most commonly they are used to prevent spammers from obtaining email addresses with systems like SpamFreeze, or prevent robots from filling in any number of online forms.

The test could be in any number of different formats. Usually they require that a human somehow identify a jumbled word, picture or sound that would be difficult for a machine to. For example, some tests require that users distinguish pictures of cats from bunnies.

reCAPTCHA is a novel system that combines the need to digitize old books and prevent spam email and other mechanized online abuse.

Older books may have type that's faded or otherwise become difficult to recognize over the years. While a person can still read them, OCR software has significant difficulties with the degraded type.

reCAPTCHA presents the user with two words that they are required to recognize to pass the test. One word is a traditional CAPTCHA, the other word is actually from a book in the process of being digitized. If the user gets the first word right, odds are they'll provide the correct answer for the word going through optical character recognition. This information is used to improve the reliability of the OCR process.

reCAPTCHA's Mailhide project uses this technology and is highly similar to SpamButcher's SpamFreeze.

Back