Knowledge domain or region specific "CAPTCHA" human/bot tests

Since generic CAPTCHA type human challenges are failing I wonder if anyone is making use of domain specific knowledge or locale specific knowledge in their bot challenges? For example, instead of the arithmetic challenge on my blog, I would ask a Sun-specific question such as:

  • What popular VM-based programming language was developed by Sun?
  • How many threads are in a T2000 core?
Whereas a blog or newspaper website based here in Dublin might separate humans from bots by challenging them on local knowledge:
  • Which Dublin stadium was the venue for the Special Olympics in 2003?
  • Which famous writer once lived in the Martello tower in Sandycove?
  • Is it possible to take a train from Cork to Sligo without going through Dublin?

Does anyone use this technique? Generic CAPTCHA challenges eventually become economical to break via computerized pattern recognition techniques or Amazon's "Mechanical Turk." A CAPTCHA at a small town English language newspaper might be broken in order to promote a scam website halfway around the world because the existence of the SPAM links improve the page rank of the scam website. However knowledge domain and region specific challenges would require that the spammer research knowledge domains and specific "local knowledge" of a region. This wouldn't scale well and the economics might not favor the spammer.

Comments:

Post a Comment:
Comments are closed for this entry.
About

bnitz

Search

Archives
« April 2014
MonTueWedThuFriSatSun
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
    
       
Today