Rather than trying to discover if a user is human, focus on the
opposite. Look for inhuman actions. The honeypot option is a good
initial barrier. But monitor that user's actions from that point on.
Are they doing something at super-human speed? Are then repeating
themselves a lot? Repeating ones self doesn't just mean the same
text exactly, it might mean almost the same text.

Perhaps it is valid for a user to sometimes include a volume of
gibberish, but scan all posts for consistent gibberish.

These are just a few things you can do to avoid bothering the humans.

There are undoubtedly many more and better ways to discover someone
as a computer.

Above all, take the penalty-box angle on all suspicions of being
inhuman. First, 1 min. Then 5. Then 30. Then an hour. Then a day. No
need to go longer than a day, really.


. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Posted from the new ixda.org
http://www.ixda.org/discuss?post=43847


________________________________________________________________
Welcome to the Interaction Design Association (IxDA)!
To post to this list ....... [email protected]
Unsubscribe ................ http://www.ixda.org/unsubscribe
List Guidelines ............ http://www.ixda.org/guidelines
List Help .................. http://www.ixda.org/help

Reply via email to