Abstract
Knowledge-based authentication systems generally rely upon users' explicit recollection of passwords, facts, or personal preferences. These systems impose a cognitive burden that often results in forgotten secrets or secrets with poor entropy. We propose an authentication system that instead draws on implicit memory-that is, the unconscious encoding and usage of information. In such a system, a user is initially presented with images of common objects in a casual familiarization task. When the user later authenticates, she is asked to perform a task involving a set of degraded images, some of which are based upon the images in the familiarization task. The prior exposure to those images influences the user's responses in the task, thereby eliciting authentication information. We ran a user study to investigate the plausibility of our system design. Our results suggest that implicit memory has potential as a basis for low-cognitive-overhead, high-stability, knowledge-based authentication. Copyright 2011 ACM.
Original language | English |
---|---|
Title of host publication | CHI 2011 - 29th Annual CHI Conference on Human Factors in Computing Systems, Conference Proceedings and Extended Abstracts |
Publisher | Association for Computing Machinery |
Pages | 2615-2618 |
ISBN (Print) | 9781450302289 |
DOIs | |
Publication status | Published - 2011 |
Externally published | Yes |