reCAPTCHA: Human-Based Character Recognition via Web Security Measures

@article{citeulike:3125452, abstract = {CAPTCHAs (Completely Automated Public Turing test to tell Computers and Humans Apart) are widespread security measures in the World Wide Web that prevent automated programs from abusing online services. They do so by asking humans to perform a task that computers cannot yet perform, such as deciphering distorted characters. Our research explored whether such human effort can be channeled into a useful purpose: helping to digitize old printed material by asking users to decipher scanned words from books that computerized optical character recognition (OCR) failed to recognize. We showed that this method can transcribe text with word accuracy over 99\%, matching the guarantee of professional human transcribers. Our apparatus is deployed in over 40,000 Web sites and has transcribed over 440 million words. 10.1126/science.1160379}, author = {von Ahn, Luis and Maurer, Benjamin and Mcmillen, Colin and Abraham, David and Blum, Manuel}, citeulike-article-id = {3125452}, doi = {10.1126/science.1160379}, journal = {Science}, keywords = {books, captcha, scanner, technology, webdesign}, month = {August}, pages = {1160379+}, posted-at = {2008-08-15 11:11:31}, priority = {2}, title = {reCAPTCHA: Human-Based Character Recognition via Web Security Measures}, url = {http://dx.doi.org/10.1126/science.1160379}, year = {2008} }

TY - JOUR ID - citeulike:3125452 L3 - citeulike-article-id:3125452 N2 - CAPTCHAs (Completely Automated Public Turing test to tell Computers and Humans Apart) are widespread security measures in the World Wide Web that prevent automated programs from abusing online services. They do so by asking humans to perform a task that computers cannot yet perform, such as deciphering distorted characters. Our research explored whether such human effort can be channeled into a useful purpose: helping to digitize old printed material by asking users to decipher scanned words from books that computerized optical character recognition (OCR) failed to recognize. We showed that this method can transcribe text with word accuracy over 99%, matching the guarantee of professional human transcribers. Our apparatus is deployed in over 40,000 Web sites and has transcribed over 440 million words. 10.1126/science.1160379 JF - Science TI - reCAPTCHA: Human-Based Character Recognition via Web Security Measures SP - 1160379 KW - books KW - captcha KW - scanner KW - technology KW - webdesign AU - von Ahn, Luis AU - Maurer, Benjamin AU - Mcmillen, Colin AU - Abraham, David AU - Blum, Manuel PY - 2008/08/14/ UR - http://dx.doi.org/10.1126/science.1160379 ER -