Friday, May 25, 2007
Everyday 150,000 hours are spent by people solving those jumpled text puzzles to register at websites or place orders online. This made Luis von Ahn say "hmm". Luis is an assitant professor of computer science at Carnegie Mellon. His idea to better use those 150,000 hours? Help digitize books, of course! Rather than figuring out which letters and numbers you are supposed to type, researchers have developed a way for people to type snippets of books. This will still confirm that we aren't robots, and also speed up the process of getting searchable texts online.
Usually books are scanned, using Optical character recognition to digitize the texts. But von Ahn said OCR doesn't always work on text that is older, faded or distorted. In those cases, often the only way to digitize the works is to manually type them into a computer.
So, the Ineternet Archive (which scans 12 thousand books every month) would send Luis hundreds of thousands of scans that the computer doesn't recognize. Those files are placed onto von Ahn's server and split up into single words that can be used as CAPTCHAs at sites all over the Internet.
Pretty cool right?
Read Story.
Related Links:
http://www.recaptcha.net
http://www.gutenberg.org
Labels: books, internet, technology
the archive
support a cause
the latest reads
search the site
get your geek fix