Continual retraining of keystroke dynamics based authenticator

Pilsung Kang, Seong Seob Hwang, Sungzoon Cho

Research output: Chapter in Book/Report/Conference proceedingConference contribution

51 Citations (Scopus)

Abstract

Keystroke dynamics based authentication (KDA) verifies a user based on the typing pattern. During enroll, a few typing patterns are provided, which are then used to train a classifier. The typing style of a user is not expected to change. However, sometimes it does change, resulting in a high false reject. In order to achieve a better authentication performance, we propose to continually retrain classifiers with recent login typing patterns by updating the training data set. There are two ways to update it. The moving window uses a fixed number of most recent patterns while the growing window uses all the new patterns as well as the original enroll patterns. We applied the proposed method to the real data set involving 21 users. The experimental results show that both the moving window and the growing window approach outperform the fixed window approach, which does not retrain a classifier.

Original languageEnglish
Title of host publicationAdvances in Biometrics - International Conference, ICB 2007, Proceedings
Pages1203-1211
Number of pages9
Publication statusPublished - 2007 Dec 1
Externally publishedYes
Event2007 International Conference on Advances in Biometrics, ICB 2007 - Seoul, Korea, Republic of
Duration: 2007 Aug 272007 Aug 29

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume4642 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other2007 International Conference on Advances in Biometrics, ICB 2007
CountryKorea, Republic of
CitySeoul
Period07/8/2707/8/29

Fingerprint

Classifiers
Authentication
Classifier
Updating
Datasets
Update
Verify
Experimental Results

ASJC Scopus subject areas

  • Biochemistry, Genetics and Molecular Biology(all)
  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Kang, P., Hwang, S. S., & Cho, S. (2007). Continual retraining of keystroke dynamics based authenticator. In Advances in Biometrics - International Conference, ICB 2007, Proceedings (pp. 1203-1211). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4642 LNCS).

Continual retraining of keystroke dynamics based authenticator. / Kang, Pilsung; Hwang, Seong Seob; Cho, Sungzoon.

Advances in Biometrics - International Conference, ICB 2007, Proceedings. 2007. p. 1203-1211 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4642 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Kang, P, Hwang, SS & Cho, S 2007, Continual retraining of keystroke dynamics based authenticator. in Advances in Biometrics - International Conference, ICB 2007, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 4642 LNCS, pp. 1203-1211, 2007 International Conference on Advances in Biometrics, ICB 2007, Seoul, Korea, Republic of, 07/8/27.
Kang P, Hwang SS, Cho S. Continual retraining of keystroke dynamics based authenticator. In Advances in Biometrics - International Conference, ICB 2007, Proceedings. 2007. p. 1203-1211. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
Kang, Pilsung ; Hwang, Seong Seob ; Cho, Sungzoon. / Continual retraining of keystroke dynamics based authenticator. Advances in Biometrics - International Conference, ICB 2007, Proceedings. 2007. pp. 1203-1211 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{a5b15f4b52ce47aab6e620a171e34719,
title = "Continual retraining of keystroke dynamics based authenticator",
abstract = "Keystroke dynamics based authentication (KDA) verifies a user based on the typing pattern. During enroll, a few typing patterns are provided, which are then used to train a classifier. The typing style of a user is not expected to change. However, sometimes it does change, resulting in a high false reject. In order to achieve a better authentication performance, we propose to continually retrain classifiers with recent login typing patterns by updating the training data set. There are two ways to update it. The moving window uses a fixed number of most recent patterns while the growing window uses all the new patterns as well as the original enroll patterns. We applied the proposed method to the real data set involving 21 users. The experimental results show that both the moving window and the growing window approach outperform the fixed window approach, which does not retrain a classifier.",
author = "Pilsung Kang and Hwang, {Seong Seob} and Sungzoon Cho",
year = "2007",
month = "12",
day = "1",
language = "English",
isbn = "9783540745488",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "1203--1211",
booktitle = "Advances in Biometrics - International Conference, ICB 2007, Proceedings",

}

TY - GEN

T1 - Continual retraining of keystroke dynamics based authenticator

AU - Kang, Pilsung

AU - Hwang, Seong Seob

AU - Cho, Sungzoon

PY - 2007/12/1

Y1 - 2007/12/1

N2 - Keystroke dynamics based authentication (KDA) verifies a user based on the typing pattern. During enroll, a few typing patterns are provided, which are then used to train a classifier. The typing style of a user is not expected to change. However, sometimes it does change, resulting in a high false reject. In order to achieve a better authentication performance, we propose to continually retrain classifiers with recent login typing patterns by updating the training data set. There are two ways to update it. The moving window uses a fixed number of most recent patterns while the growing window uses all the new patterns as well as the original enroll patterns. We applied the proposed method to the real data set involving 21 users. The experimental results show that both the moving window and the growing window approach outperform the fixed window approach, which does not retrain a classifier.

AB - Keystroke dynamics based authentication (KDA) verifies a user based on the typing pattern. During enroll, a few typing patterns are provided, which are then used to train a classifier. The typing style of a user is not expected to change. However, sometimes it does change, resulting in a high false reject. In order to achieve a better authentication performance, we propose to continually retrain classifiers with recent login typing patterns by updating the training data set. There are two ways to update it. The moving window uses a fixed number of most recent patterns while the growing window uses all the new patterns as well as the original enroll patterns. We applied the proposed method to the real data set involving 21 users. The experimental results show that both the moving window and the growing window approach outperform the fixed window approach, which does not retrain a classifier.

UR - http://www.scopus.com/inward/record.url?scp=37849027593&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=37849027593&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:37849027593

SN - 9783540745488

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 1203

EP - 1211

BT - Advances in Biometrics - International Conference, ICB 2007, Proceedings

ER -