There seems to be a lack of data in assessing how the cognitive load of using ever more complex passwords can be reduced by employing an interactive human-machine exchange for the development of more "natural" passwords and access codes.
Can you please provide a short description of what SNAP is? I tried skimming your cited article but it is rather thick reading. From what I understood SNAP tries to increase the password space by associating a password key with a conditioned response in users?
The CUPS lab at Carnegie Mellon University has been looking at the effects of memory and passwords. Serge Engelman at Berkley has also been looking at passwords and challenge questions. (Disclaimer I used to work with Serge in the CUPS lab.)
There was some work done quite a while ago looking at having humans and computers complete an exchange via a mathematical formula. Unsurprisingly humans aren't very good at math so this didn't go far.
You might actually want to look into research on out of band two factor authentication. There have been several papers, including some at CHI and SOUPS this year, which looked at ways that a user could verify themselves in a non-obvious way. One example being having a unique code sent to their cell phone. Or having tactile feedback given to the user which they then use to construct a unique password each time (CHI 2010?).
Alternatively there has been work by Roy Maxion at Carnegie Mellon on the use to biometrics for continuous authentication. In particular he has been looking at the ways people type on keyboards which can continuously be used to authenticate them and is harder to replicate than a fingerprint.
There has been piles of research into graphical type passwords. Mike Reiter did research on having users construct stories and then select the person, place, and thing associated with their story as their password. This did not work well. SOUPS has had several papers on how people select graphical passwords and the general conclusion is that, depending on the image, most people select the same points on the image. There was also some research out of the CUPS lab (SOUPS 2013?) looking at using easier to remember passwords that used full words instead of letters so you just had to remember a set of words.
I haven't seen any research specifically examining cognitive load in conjunction with passwords, though I would not be surprised if it exists.
The notion of SNAP has to do with understanding that all human-machine interactivity is necessarily framed in time. It is time that is both the qualitative and quantitative essence of all "real-time" algorithms in general, and "real-time" metrics such as the articulated passwords that typify the Symbolic/Non-symbolic Access Protocol. As such, time is the defining parameter of all human-machine interfaces.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Human-Computer Interaction