Commonsense Captcha: Difference between revisions
(small changes) |
m (qualified bold claim) |
||
Line 14: | Line 14: | ||
website. This project uses a new type of Captcha that asks a series of | website. This project uses a new type of Captcha that asks a series of | ||
commonsense questions. This new Captcha is accessible to vision | commonsense questions. This new Captcha is accessible to vision | ||
impaired users and is able to distinguish between human and computers | impaired users and (time will tell) is able to distinguish between human and computers | ||
effectively. Most importantly, it allows individuals to, in the | effectively. Most importantly, it allows individuals to, in the | ||
process of their normal actions, make easy and constructive | process of their normal actions, make easy and constructive | ||
contributions to AI research by helping verify and shared a | contributions to AI research by helping verify and shared a | ||
commonsense databases. | commonsense databases. |
Revision as of 00:46, 3 January 2006
The original proposal (now a little bit dated as it was written before the project was implemented or even fully designed) is over at Benjamin Mako Hill's personal wiki at: http://wiki.mako.cc/CommonSenseApproachToSpam
Short Description
A Captcha is a "Completely Automated Public Turing Test to tell Computers and Humans Apart." Captchas are one tool frequently used to stop spammers from using computer programs to register for web-accounts or to deface wiki pages with advertisements. Captcha's are most familiar as a picture of garbeled or distorted text that humans are asked to decode on a website. This project uses a new type of Captcha that asks a series of commonsense questions. This new Captcha is accessible to vision impaired users and (time will tell) is able to distinguish between human and computers effectively. Most importantly, it allows individuals to, in the process of their normal actions, make easy and constructive contributions to AI research by helping verify and shared a commonsense databases.