Touchpad Input for Continuous Biometric Authentication

Abstract : Authentication is a process which is used for access control in computer security. However, common existing methods of authentication, which are based on authentication during the login stage, are insecure due to the lack of authentication after the initial instance. Ideally, authentication should be continuous and should not interfere with a user’s normal behavior as to not create an inconvenience for the user. Behaviometric identification, for example, verifies a user’s identity based on his behavior, both continuously and without interruption. This work shows that it is possible, with great accuracy, to identify different users based on their touchpad behaviors. While linear classifiers proved ineffective at classifying touchpad behavior, kernel density estimation and decision tree classification each proved capable of classifying data sets with over 90% accuracy.
Complete list of metadatas

Cited literature [11 references]  Display  Hide  Download

https://hal.inria.fr/hal-01404189
Contributor : Hal Ifip <>
Submitted on : Monday, November 28, 2016 - 2:46:53 PM
Last modification on : Tuesday, August 13, 2019 - 1:42:01 PM
Long-term archiving on : Monday, March 20, 2017 - 5:01:13 PM

File

978-3-662-44885-4_7_Chapter.pd...
Files produced by the author(s)

Licence


Distributed under a Creative Commons Attribution 4.0 International License

Identifiers

Citation

Alexander Chan, Tzipora Halevi, Nasir Memon. Touchpad Input for Continuous Biometric Authentication. 15th IFIP International Conference on Communications and Multimedia Security (CMS), Sep 2014, Aveiro, Portugal. pp.86-91, ⟨10.1007/978-3-662-44885-4_7⟩. ⟨hal-01404189⟩

Share

Metrics

Record views

105

Files downloads

326