Web Authentication Using Mikon Images

Abstract

Authentication is mostly achieved by means of the ubiquitous password. This is suboptimal in some settings, such as for user groups with cognitive or language difficulties. Many web-based systems have user groups with widely ranging capabilities, and more innovative authentication mechanisms should be investigated to enhance usability and accessibility while still delivering the required level of security to authorise legitimate users.This paper presents details of an authentication system which relies on the user identifying previously drawn Mikons 1 . Mikons are self-drawn icon-like images, meant to depict a message the artist wants to convey at that point in time. These are drawn, at enrolment, using an embedded ShockWave component within a browser. At authentication the user identifies his or her own Mikons from challenge sets, each containing one of the user's Mikon and a number of distractor Mikons.The efficacy of Mikons in this setting was investigated by using them in a recognition-based authentication system to authorise users of an online homework system over an eight month period. The Mikon-based system performed very well in terms of memorability and scalability, as anticipated, thus achieving the level of accessibility hoped for. A measure of predictability was observed, with a few of the participants being able to link sets of Mikons to their creators; but this did not pose a security risk to the system. This study shows that Mikon authentication has the potential to be a viable alternative to passwords for systems where the security requirement is secondary to other, more important, considerations. Such systems are usually low-risk and are often used by users with developmental, language or cognitive difficulties, or by users who are not yet literate. The imposition of a password on such users can be overly stringent and excessively demanding in terms of scarce cognitive resources. In this context, therefore, Mikons are a viable alternative to meet the needs of the target user group.

A Comprehensive Study of the Usability of Multiple Graphical Passwords

“…The idea is that humans can remember images better than recalling alphanumeric text [4,5] and so this may be a way of devising more memorable passwords. Given the need for alternative usable authentication systems and existing interest in image passwords [3,6,7,8,9,10] as a potential solution, we identify an important limitation of existing work: most prior studies with images as passwords, except [11,12], have focused on the usability of a single password. However, people will need to remember and use multiple image passwords in the same way that they currently use multiple alphanumeric passwords.…”

“…Existing studies have reported the usability of faces [3,7], Mikons [6,10], doodles [6,9], abstract art [8] and pictures of daily objects [7]; however, these studies have focused only on the use of single passwords. We are aware of only two pieces of prior work that studied multiple image passwords: [12] studied the use of multiple facial passwords and [11] compared multiple picture passwords to that of multiple PIN's.…”

“…A three stage study was designed, which consisted of: (1) a prestudy questionnaire evaluating the participant demographics and current password strategies; (2) an 8-week online study of participants using multiple image passwords; and, (3) a post-study questionnaire regarding participant experiences. In our usability study we used the following image types (See Appendix): (1) Mikon: These are iconlike images which have been drawn by users in studies reported in [6,10] using a tool called the Mikon engine developed by Mikons.com. (2) Doodle: These images are drawn by a user using pen on paper, in studies reported in [6,9].…”

An Algorithm for Automatically Choosing Distractors for Recognition Based Authentication using Minimal Image Types

“…These images have superior memorability because they have been provided by the users. As explained in [30], the actions engaged in during production of these images enhance memorability.…”

“…Different image types have been trialled, with some demonstrating more efficacy than others: faces [10], abstract art [25], system provided or user-taken photographs [26,27], icons [28], doodles [29] and Mikons [30]. A comparison between photographs, user photographs and doodles showed that doodles performed best in terms of memorability [31].…”