• E-mail this page
• |  Print this page
• | 
• | 
• | 

Google Revisits Crowdsourcing With reCAPTCHA Acquisition

In its second acquisition this year, Google has bought reCAPTCHA, a company that provides CAPTCHA images as a barrier to online fraud.

The term CAPTCHA is an acronym for the phrase "Completely Automated Public Turing test to tell Computers and Humans Apart."

Luis von Ahn, co-founder of reCAPTCHA, was among the computer scientists from Carnegie Mellon and IBM who coined the term back in 2000.

What makes reCAPTCHA interesting for Google is that it kills two birds with one stone. Not only are CAPTCHAs necessary for online security -- to prevent spammers from using scripts to automatically register thousands of Gmail accounts, for example -- but reCAPTCHA's unique technology is designed to cull the phrases it presents to users from scanned books.

Because of this, Google will be able to improve the accuracy of the optical character recognition (OCR) applied to book scans through what amounts to "crowdsourced" copy editing. In so doing, Google is again finding value in aggregated intelligence: The highly relevant search results that made Google's name owe a lot to the PageRank algorithm developed by company co-founders Sergey Brin and Larry Page. PageRank weighs links between Web pages as if they were votes for relevance, thereby leveraging the judgment of the crowd to determine which Web sites matter.

"reCAPTCHA's unique technology improves the process that converts scanned images into plain text, known as Optical Character Recognition," explain von Ahn, co-founder of reCAPTCHA, and Will Cathcart, a Google product manager, in a blog post. "This technology also powers large scale text scanning projects like Google Books and Google News Archive Search. Having the text version of documents is important because plain text can be searched, easily rendered on mobile devices and displayed to visually impaired users. So we'll be applying the technology within Google not only to increase fraud and spam protection for Google products but also to improve our books and newspaper scanning process."

Google has no shortage of errors to correct. One of the company's Book Search engineers recently acknowledged that there are millions of errors in the metadata used to describe the books scanned for Google Book Search. No doubt the company's OCR output isn't perfect either.

But such problems look a lot less daunting when one can leverage CAPTCHA input to correct errors.

« Whatever Happened To The Idea Of Instant On? | Main | Health IT Is Part Of Senate Committee's New Healthcare Reform Bill »