| With a new Terminator movie coming out, I started thinking about the ways that on the Internet now we need to try and distinguish the humans from the robots. There is already a “war” going on out there between humans and sophisticated robots for control of Facebook, Hotmail, Gmail and other web 2.0 services. |  |
One of the more annoying “fronts” in this war is CAPTCHA (“Completely Automated Public Turing Test to Tell Computers and Humans Apart”). This reverse Turing test is what is used on many sites to test to see if you are human and not some automated robot.
The problem is that these tests are like antibiotics – the more they get used the more sophisticated the robots. Early CAPTCHA tests were easily broken by bots so they have evolved into more and more sophisticated tests. Now the number of hoops and the readability of these tests are so bad that human beings can barely pass them.
The challenge is how to make a test easily answerable by a human being but yet heuristically difficult enough to foil a software program.
Google is experimenting with new techniques to try and simplify the experience of verification – see below.
Even using this technique, Google has to specifically pick the right images:
However, while the images will be randomly rotated, they will be carefully selected. First, Google will exclude images for which its own computers can readily identify the top, such as photographs showing landscapes with blue sky (easily detected), text (easily recognized) or portraits of people (there are many facial recognition applications on the market). Then, it will screen out images that humans find too difficult to orient (for example, abstract art, or overhead views, that don't have a readily identifiable top) by conducting a sort of opinion poll.
Google is relying not just on individual results but collective results. The heuristic picture composed from millions of clicks is what will power the analytics so that Google can reliably detect bots instead of humans.
No comments:
Post a Comment