Half a croissant, on a plate, with a sign in front of it saying '50c'
h a l f b a k e r y
Number one on the no-fly list

idea: add, search, annotate, link, view, overview, recent, by name, random

meta: news, help, about, links, report a problem

account: browse anonymously, or get an account and write.




Agent monitoring keyboard input could identify user from typing patterns
  (+3, -1)
(+3, -1)
  [vote for,

How many times do you make the same typing errors? My initials DJA I type thousands of times a day but for some reason i always type d-a-j-backspace-backspace-j-a. I can do this instantly and i dont even think but i'm willing to bet that should this sequence be monitored in terms of time taken to type and time between key depressions - no one on earth could type it in *exactly* the same way as I do. Now extrapolate - monitor my typing for say 10,000 words and record every single keystroke including errors and times (add a keyboard using velocity and you could monitor the force with which i hit the keys) and i suspect a very accurate and very unique "keyboardprint" could be produced.

In these security conscious times this could be the next best thing since passwords. Time and time again, we are told that the simplest and most effective form of security is a good password. If someone has the appropriate password for whatever they are accessing (and usually Username) then most security systems will take no further action. My idea is the next stage. Assuming that the user has logged on successfully, the computer can then monitor a set amount of typing (obviously room for calibration here) and check that the typists keystrokes match the previously recorded KeyboardPrint. If they dont - disconnected/logged out whatever.

I realise this is fairly heavy handed but in extremely secure scenarios this could be a good option.

dja, Oct 31 2000

Typing Patterns for Authentication http://it.slashdot....2328204&threshold=1
slashdot discussion [xaviergisz, Apr 20 2007]


       Because fingerprint recognition requires ridiculous, expensive hardware. KeyboardPrint (or BioPassword) can be done purely in software.
egnor, Oct 31 2000

       My boss' Mercedes learns his driving style. If the car is driven by someone who doesn't accelerate, brake, turn, etc. the way he does, it asks the driver to enter a security code. If the code is not correct, the car shuts down (and, I believe, calls the police).   

       I would think typing patterns would be more reliable for identification than driving patterns. But you would still need a PIN override for days when you're in a hurry, on medication, etc.
beauxeault, Oct 31 2000

       I thought of the fingerprint thing but this would then mean that it was implemented on every keyboard which is impractical. My system would work remotely, locally whatever.
dja, Nov 01 2000

       I think I'v heard of this before, maybe in Neal Stephensons "cryptonomicon". I thought it was a good idea.
woodja, Nov 02 2000

       Hell, if it were -really- good, it could identify particular individual CATS.
Gee, that's useful.

       Oh, I doubt that it could work remotely, because of lag - even if the lag is a few milliseconds, that's the sort of precision you're talking about. Unless this system is built into the keyboard itself, somehow, but that would be exactly what you were talking about.   

       What would be interesting is if this software was modified so that, say, if a person is holding a gun to your head, you hit the keys harder and slower (deliberately, but the gun-holder won't know that), and the computer pretends there was a power cut or something... Oh, I don't know.
Detly, Nov 04 2000

       Interesting idea; when we can perfect this software and validate it it would definitely be of use.
sike, Aug 22 2003


back: main index

business  computer  culture  fashion  food  halfbakery  home  other  product  public  science  sport  vehicle