Halt! Be Ye Human or Bot?
There are many times when a website needs to collect data from the visitor: to sign up for membership, to send an email to the website
administrator(s), to make a purchase, to add entries in a blog, and so on. Web 2.0 wants every body to be able to do these things. Part
of the idea of Web 2.0 is that the internet will continue to be created by website administrators AND the visitors that use the sites,
by adding content of their own. This means that increasing amounts of content will come from your members and other visitors to your website.
The problem with this is the high (and increasing) number of hackers and cyber-terrorists who are always ready to take advantage of the systems out there, including your own website. People with malicious intent on the internet create software to accomplish their ends, since, once set in motion, software will continue its attack relentlessly, against 1,000's or more websites a day. You want your visitors to be able to send email to you and to other members of your website, to post to your blog(s) and to signup for membership. But you don't want malicious software coming in and hi-jacking your systems for its own use. WebForm Defender is an easy to use system meant to help secure your web forms from typical abuse. The basic idea is this: how can you tell a "user" is really a person and not a piece of software (a data-miner, a bot, a spider, or other general malware)? This is the essential question. In principal, this test is simple... |
C.A.P.T.C.H.A.
Completely Automated Public Turing test to tell Computers and Humans Apart
Alan Turing (shown at left), in all ways considered the father of modern computing, made a rather remarkable contribution
to the science of artificial intelligence. While thinkers of the time were getting dizzy over debates about whether a machine
could ever be "conscious" or "self aware," Turing side-stepped the issue brilliantly by saying that, since it could probably never be proven one way or the other, self-awareness in a computer was not a valid goal for artificial intelligence.
What did matter, he said,was whether a machine could fool you into thinking it was human by its actions. If, for example, a program were
to talk with you over chat, IM or text-message for a period of time, say 20 minutes, and you still believe you are talking to a person, then it has
passed the Turing Test.
CAPTCHA is a similar idea, geared toward separating out humans from software. There are things that a human can do that a computer (so far) is simply no good at. One of these things is "separating data from noise" -- such as taking a distorted image reading it correctly, or understanding a spoken phrase against a lot of background noise.
