Shared dataset on natural human-computer interaction to support continuous authentication research

Murphy, Christopher; Huang, Jiaju; Hou, Daqing; Schuckers, Stephanie

doi:10.1109/btas.2017.8272738

Cited by 35 publications

(26 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The performance of the G&P approach on the Clarkson dataset is much worse than the 10.4% reported in (Murphy et al 2017) because 10,000 keystroke events were used for enrollments and 1,000 keystroke events were used as samples for test vectors in that study. This and G&P's inability to use small data samples demonstrates the data efficiency of D-Vectors, in that it can achieve similar or greater performance using far less data for enrollment and testing.…”

Section: Keystroke Verification Resultsmentioning

confidence: 71%

“…The first dataset (MultiMod) is one we collected; it contains both keystroke and gait data, which we describe in detail below 1 . Additionally, we use two publicly available datasets as benchmarks -one for keystroke (Clarkson) (Murphy et al 2017) and gait (UCI) (Anguita et al 2013).…”

Section: Datasetsmentioning

confidence: 99%

“…The Clarkson benchmark keystroke dataset (Murphy et al 2017) consists of 103 users collected over 2.5 years from subjects' personal machines during normal interactions, representing a realistic use case. The dataset consists of 87 users with sufficient data.…”

Section: Datasetsmentioning

confidence: 99%

“…These shortcomings have been demonstrated on several occasions by keystroke dynamics researchers. Killourhy & Maxion (Killourhy and Maxion 2009) and Murphy et al (Murphy et al 2017) compared algorithms reported by several different studies on various datasets, including real-world datasets, and found that the performance of these algorithms often degraded significantly outside of laboratory settings.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Beyond Speech: Generalizing D-Vectors for Biometric Verification

Baldwin

Burnham²,

Meyer

et al. 2019

AAAI

View full text Add to dashboard Cite

Deep learning based automatic feature extraction methods have radically transformed speaker identification and facial recognition. Current approaches are typically specialized for individual domains, such as Deep Vectors (D-Vectors) for speaker identification. We provide two distinct contributions: a generalized framework for biometric verification inspired by D-Vectors and novel models that outperform current stateof-the-art approaches. Our approach supports substitution of various feature extraction models and improves the robustness of verification tests across domains. We demonstrate the framework and models for two different behavioral biometric verification problems: keystroke and mobile gait. We present a comprehensive empirical analysis comparing our framework to the state-of-the-art in both domains. Our models perform verification with higher accuracy using orders of magnitude less data than state-of-the-art approaches in both domains. We believe that the combination of high accuracy and practical data requirements will enable application of behavioral biometric models outside of the laboratory in support of much-needed improvements to cyber security.

show abstract

Section: Keystroke Verification Resultsmentioning

confidence: 71%

Section: Datasetsmentioning

confidence: 99%

Section: Datasetsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Beyond Speech: Generalizing D-Vectors for Biometric Verification

Baldwin

Burnham²,

Meyer

et al. 2019

AAAI

View full text Add to dashboard Cite

show abstract

“…Otherwise, in real-world scenarios, users usually type only a few characters to search or chat online and spend the rest of the time browsing the web, documents, pictures, videos, etc. We find that the number of keystrokes less than 50 in once typing accounts for 51.95% of all keystrokes in the Clarkson II dataset (Murphy et al 2017). It is necessary to overcome these difficulties of using short keystroke data for accurate KCA in real uncontrolled scenarios.…”

Section: Introductionmentioning

confidence: 96%

TKCA: a timely keystroke-based continuous user authentication with short keystroke sequence in uncontrolled settings

Yang

You

et al. 2021

Cybersecur

View full text Add to dashboard Cite

Keystroke-based behavioral biometrics have been proven effective for continuous user authentication. Current state-of-the-art algorithms have achieved outstanding results in long text or short text collected by doing some tasks. It remains a considerable challenge to authenticate users continuously and accurately with short keystroke inputs collected in uncontrolled settings. In this work, we propose a Timely Keystroke-based method for Continuous user Authentication, named TKCA. It integrates the key name and two kinds of timing features through an embedding mechanism. And it captures the relationship between context keystrokes by the Bidirectional Long Short-Term Memory (Bi-LSTM) network. We conduct a series of experiments to validate it on a public dataset - the Clarkson II dataset collected in a completely uncontrolled and natural setting. Experiment results show that the proposed TKCA achieves state-of-the-art performance with 8.28% of EER when using only 30 keystrokes and 2.78% of EER when using 190 keystrokes.

show abstract

Performance Analysis of Free Text Keystroke Authentication Using XGBoost

Kuzminykh,

Mathur,

Ghita

2023

Lecture Notes on Data Engineering and Communications Technologies

View full text Add to dashboard Cite

Authentication based on keystroke dynamics is a form of behavioral biometric authentication that uses the user typing patterns and keyboard interaction as a discriminatory input. This type of authentication can be coupled with a fixed text password in a traditional login system to contribute to a multifactor authentication or provide continuous user authentication in a usable security system, where the typing patterns are continuously analysed to validate the user at run time. This paper investigates the effectiveness of free text keystroke for continuous authentication in real-world systems. Evaluation is performed using XGBoost multiclass classification, applied to an unbalanced free-text keystroke dataset. The introduction of additional activity-based features and removal of inaccuracies in the timing between keys allowed a reduction of the EER for the Clarkson II dataset from 14-24%, as achieved by previous studies, to 8% when employing the proposed method.

show abstract

Shared dataset on natural human-computer interaction to support continuous authentication research

Cited by 35 publications

References 15 publications

Beyond Speech: Generalizing D-Vectors for Biometric Verification

Beyond Speech: Generalizing D-Vectors for Biometric Verification

TKCA: a timely keystroke-based continuous user authentication with short keystroke sequence in uncontrolled settings

Performance Analysis of Free Text Keystroke Authentication Using XGBoost

Contact Info

Product

Resources

About