On the Internet, nobody knows you’re a dog.
in The New Yorker, 1993
It may be OK to be a dog, but these days more and more sites ask you to prove you’re not a computer program (aka a “bot”). Bots wreck havoc around the net, signing up for mailing lists (so they can send spam), commenting on weblogs (with thinly disguised spam), stuffing ballot boxes, and “scraping” pages for anything they can get (including email addresses so they can send more spam). More and more sites are forcing human visitors to prove they are human by solving a “simple” puzzle called a Captcha
.
The term Captcha was coined in 2000 by researchers from Carnegie Mellon University and IBM, although similar ideas were apparently in use some years prior to 2000. The name is a cutesy pseudo-acronym built from the descriptive phrase “Completely Automated Public Turing test to tell Computers and Humans Apart”. (A “Turing test” is, in itself, supposed to be a way to tell computers and humans apart, but I digress…).
The most common type of captcha requires that the user type the contents of a distorted or otherwise obfuscated sequence of letters or digits that appears on the screen.
Unfortunately, depending on the level of obfuscation, I tend to “fail” these tests perhaps 10-20% of the time. Capital B’s look too much like 8’s. Is that a 7 or an angled 1? A 6 or a G? What’s that, a letter or a number? Can I try again?
I’ve seen occasional variations on the theme. One site I visited provided a small arithmetic quiz, e.g.
5 + 7 =
I liked that idea. Here’s a variant, phrased as a “word problem”.
“What is fifty times ‘a’, if ‘a’ is three?”
Or, how about this one, which uses CSS and JavaScript on the client combined with a back-end script to create the captcha and process what the visitor types into the box.
Of course, it might not work if you’re color-blind… (but a common argument against captchas is that they don’t work for the sight-impaired visitor).
Captchas can be done with pictures instead of letters and digits (although I’ve never been to a site that uses any of these in action). I like bongo, myself.
Bongo is a program that asks the user to solve a visual pattern recognition problem. In particular, Bongo displays two series of blocks, the left and the right series. The blocks in the left series differ from those in the right, and the user must find the characteristic that sets the two series apart. A possible left and right series are shown below:
Then there’s this (possibly facetious) recommendation for an intelligent captcha test, or “captcha++”.
Already, a “captcha” test is a common way of determining if the poster is a human or a machine. I think it’s time to raise the bar and determine if the poster is an intelligent human or a machine.
The standards of intelligence should be modest. We examined the freshman curricula for a degree in “Liberal Arts” from a local community college to come up with a subject set. Our questions are not meant to be the least bit tricky; in fact they should be extremely easy for any adult who paid attention in high school or attended some college courses. What it will pick out, we hope, are the people lacking a sufficient foundation to be able to contribute to a civilized discourse.
Unfortunately, this test suffers, as do all “intelligence” tests, from the requirement for the creator and the taker to share a similar background, education, and culture. I’m a well-educated member of Mensa, over the age of 40. But my educational background is in Biology and Chemistry. I never much cared for geography, social studies, or politics. I’m not a Modern Art fan. And, I don’t watch television. Consequently, I did rather poorly on this “test”.
Nevertheless, I approve of the idea behind a more intelligent captcha test, if only because I’m getting tired of trying to read those boxes full of twisted, obscured letters and numbers.
Interesting Links
- The CAPTCHA Project (from CMU)
- OCR Research Team – breaking weak CAPTCHAs and researching better ones.
- captchaservice.org — A web service that serves up CAPTCHAs, ready to be used on your site.
- John Dale’s Blog – additional commentary and pointers
- Science News Online, Week of April 16, 2005; Vol. 167, No. 16
April 07, 2006
OMG!!! Kittens!
Here’s an idea I really like: KittenAuth.
I primarily wanted a method that didn’t make them have to decipher horrible random text from a box. With OCR constantly improving, this is inherently a failed system… Its more a case of “when”, less “if”.
…
Kittens are the answer to all this. Sure you can teach a computer what a cat looks like, and it will probably have a fair shot at picking kittens from alligators but what about when you put it up against other similarly cute animals? Well… I’ve yet to have something try and bodge through it.
You can try KittenAuth here.
vlbrown-mzydt 
Leave a comment