Yesterday, I was looking for a suitable CAPTCHA module for use in one of my websites. For those of you who don’t know, CAPTCHA is an anti-bot mechanism that is used in computers for differentiating between automated programs (bots) and humans. Word verifications that you need to do before submitting most of the forms online are CAPTCHAs.
The term CAPTCHA is an acronym of “Completely Automated Public Turing test to tell Computers and Humans Apart”. A number of CAPTCHA module are available nowadays –beyond words, some ask you to solve a simple math problem and some ask you to recognize objects in a given image.
reCAPTCHA anti-bot and anti-spam mechanism is based on word verification. It presents two slightly distorted words to the user and asks her to fill those words in a text box in order to establish that the filler is a human being. This way it secures web forms from being filled in by automatic programs.
reCAPTCHA example screen
But reCAPTCHA does much more than that behind the screen. The brain behind reCAPTCHA, Luis von Ahn, realized that hundreds of thousands of hours are being spent by Internet users in solving reCAPTCHA. Daily more than 200 million CAPTCHAs are solved by human beings. So, Luis thought of channelizing this human effort for something more constructive.
Luis designed reCAPTCHA mechanism in a way so that it delivers words that remained unrecognized by Optical Character Recognition (OCR) while digitizing a book or other text. OCR failed to recognize these words but humans can definitely do a better job. Now reCAPTCHA is helping digitization of Google Books and past issues of The New York Times.
The strange thing in this scheme is this that if computer (i.e. OCR) cannot recognize the words –then how come is it going decide whether the humans have solved the reCAPTCHA correctly? The answer to this is that reCAPTCHA offers one unrecognized word and one such word whose answer is known to the computer. If the humans solve the recognized word –the system assumes that the answer to the unrecognized words was also correct. Then the system further confirms the answer by delivering the same unrecognized word to other users.
I found it quite amusing and innovative. This effort is similar to Search for Extra Terrestrial Intelligence (SETI) project in terms of maximizing output from the available resources.
Do let me know how you liked this piece of information. If you want to add anything relevant to it, please feel free to comment. Thank you for using TechWelkin.
Leave a Reply