How Will ReCAPTCHA Help Google?
Google will buy ReCAPTCHA, a company that provides CAPTCHAs to help protect more than 100,000 Web sites from spam and fraud. Financial terms were not disclosed.
CAPTCHAs (Completely Automated Public Turing Test To Tell Computers and Humans Apart) are often created from old pieces of text, including books and newspapers.
Computers, through programs called "spiders" or "robots," find it hard to recognize those words because the ink and paper have degraded over time. So far, CAPTCHA programs have been successful in deflecting robotic attacks because the spiders can't recognize the text.
But in addition, the technology can help Google in its large-scale -- and controversial -- text scanning projects such as Google Books and Google News Archive Search.
Google wants the technology because as users decipher the jumbled characters with CAPTCHAs, the software "learns" to interpret those words.
The technology that Google now uses to scan documents, Optical Character Reader Recognition, stumbles over the translation of print that's faded and worn. ReCAPTCHA's Web site illustrates that accuracy problem.
Because ReCAPTCHA uses old text from old print publications and users then type them in as a CAPTCHA, users teach computers to read the scanned text. Having the text version of documents is beneficial because it facilitates searching and renders it easily on mobile devices. ReCAPTCHA's slogan, "Stop Spam. Read Books," seems to be a good fit with Google's plan.
Therein may lie the rub, however. As we teach computers to read, CAPTCHAs may lose their appeal as a security mechanism. Of course, as hackers' malicious software also becomes more sophisticated, the distorted character strategy would become threatened anyway. Google's next step may need to address that security concern.