Downloadable Content
Download PDF
Masters Thesis
Mobile keystroke dynamics: assessment and implementation
The majority of Americans now own a smartphone. We keep our most personal data in pockets and purses, seldom prepared for the pain and distress that loss of our phones can cause. For those with high-security data on their phone, theft of the device may prove catas- trophic. Although lock screens can help prevent unauthorized access, they cannot detect unauthorized device usage. Additionally, an attacker who has learned the passcode/pattern can use the device at her discretion. This paper explores the feasibility of increasing mobile security through the applica- tion of keystroke dynamics. Keystroke dynamics is a biometric based on typing. Typists develop individualized rhythms and patterns that can be used to distinguish the authentic user from an impostor. Traditionally, keystroke dynamics has been applied in high-security applications where users are typing on a full-sized physical keyboard. With the increasing prevalence of smartphones, the application of keystroke dynamics to the mobile domain could prove a powerful weapon against mobile data theft. To this note, I have explored the mobile application and accuracy of several well-known keystroke dynamics classifiers and developed an Android Input Method that implements typing pattern recognition using the best of these, the Nearest Neighbor Mahalanobis Distance classifier.
Thumbnail | Title | Date Uploaded | Visibility | Actions |
---|---|---|---|---|
Ryan-Shea-thesis-2015.pdf | 2021-04-29 | Public | Download | |
clacker.zip | 2021-04-29 | Public | Download | |
clackerjs.zip | 2021-04-29 | Public | Download |
Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.