Tuesday 1 May 2012

reCAPTCHA - everyone writes a book

Unless you're an internet freshman, you have probably come in contact with a CAPTCHA - most likely to access some service on a website. And like me, you probably have refreshed some of them many times to generate one that's readable.

Why do we have to put up with CAPTCHA when it's so annoying? Well, as you might know, it was developed to combat the spread of automated software (bots). A CAPTCHA is a test to prove you're human - because only a human can read and interpret those scribbles.

So what exactly CAPTCHA does to be considered a human-based computation? Luis von Ahn (pictured), an early CAPTCHA developer and crowd-sourcing pioneer, has seen an opportunity to make the process useful. Enter reCAPTCHA - a system developed in Carnegie Mellon University in Pittsburgh. What makes it different? The 'scribbles' are in fact, a (distorted) scanned text - one that optical character recognition (OCR) software can't recognize. A human, however, is (usually) more perceptive, and the words that are rewritten are later turned into machine-interpretable information.

So! Next time you write a CAPTCHA, keep in mind that you're doing your part to bring a book into digital format. :)

No comments:

Post a Comment