Touchalytics: On the Applicability of Touchscreen Input as a Behavioral Biometric for Continuous Authentication
Open accessThe DOI matches its Crossref record; 29 of the 29 references it cites were found in a registry or an index.
Abstract
We investigate whether a classifier can continuously authenticate users based on the way they interact with the touchscreen of a smart phone. We propose a set of 30 behavioral touch features that can be extracted from raw touchscreen logs and demonstrate that different users populate distinct subspaces of this feature space. In a systematic experiment designed to test how this behavioral pattern exhibits consistency over time, we collected touch data from users interacting with a smart phone using basic navigation maneuvers, i.e., up–down and left–right scrolling. We propose a classification framework that learns the touch behavior of a user during an enrollment phase and is able to accept or reject the current user by monitoring interaction with the touch screen. The classifier achieves a median equal error rate of 0% for intrasession authentication, 2%–3% for intersession authentication, and below 4% when the authentication test was carried out one week after the enrollment phase. While our experimental findings disqualify this method as a standalone authentication mechanism for long-term authentication, it could be implemented as a means to extend screen-lock time or as a part of a multimodal biometric authentication system.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The cited source says what the claim attributes to it.
In the citing paper
“For example, an approach that uses the touch dynamics (e.g., [33]) for continuous authentication will not work on some other devices such as laptops or desktops.”
From Zero Trust Architecture (ZTA): A Comprehensive Survey (2022) · cites this paper as [33]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Cortes, C., & Vapnik, V. (1995) · Machine Learning · doi:10.1007/bf00994018 · Cited by 59,421
Cortes, C., & Vapnik, V. (1995) · Machine-mediated learning · doi:10.1023/a:1022627411411 · Cited by 48,445
A training algorithm for optimal margin classifiers
Boser, B. E., Guyon, I. M., & Vapnik, V. N. (1992) · Proceedings of the fifth annual workshop on Computational learning theory · doi:10.1145/130385.130401 · Cited by 11,992 · cited 1×
“In our experiments, we took 11 strokes for each decision, resulting in 11 to 43 seconds until the first decision is available.”
Method(2000) · CrossRef Listing of Deleted DOIs · doi:10.1162/153244303322753616 · Cited by 3,502
Mardia, K. V., & Jupp, P. E. (1999) · Wiley Series in Probability and Statistics · doi:10.1002/9780470316979 · Cited by 3,203
An Algorithm for Finding Best Matches in Logarithmic Expected Time
Friedman, J. H., Bentley, J. L., & Finkel, R. A. (1977) · ACM Transactions on Mathematical Software · doi:10.1145/355744.355745 · Cited by 2,813 · cited 1×
“For all consecutive decisions, a sliding window of the preceding strokes is available for robust decision.”
BackgroundBiometrics: A Tool for Information Security
Jain, A. K., Ross, A., & Pankanti, S. (2006) · IEEE Transactions on Information Forensics and Security · doi:10.1109/tifs.2006.873653 · Cited by 1,095 · cited 1×
“1 http://www.mariofrank.net/touchalytics/index.html”
Smudge attacks on smartphone touch screens
Aviv, A. J., Gibson, K. L., Mossop, E., Blaze, M., & Smith, J. M. (2010) · Cited by 659
On-line signature verification
Jain, A. K., Griess, F. D., & Connell, S. D. (2002) · Pattern Recognition · doi:10.1016/s0031-3203(01)00240-0 · Cited by 536 · cited 1×
“Our contribution is a classification framework that serves as a proof-of-concept for touch-based behavioral biometric authentication.”
MethodIdentity authentication based on keystroke latencies
Joyce, R., & Gupta, G. (1990) · Communications of the ACM · doi:10.1145/75577.75582 · Cited by 534 · cited 1×
“Furthermore, the device cannot detect intruders after the authentication step is performed successfully.”
Background
Research works printed but not linked: 8
- [34] I. Fischer, C. Kuo, L. Huang, and M. Frank, 'Smartphones: Not smart enough?' in SPSM '12: 2nd ACM CCS Workshop on Security and Privacy in Mobile Devices . New York, NY, USA: ACM, 2012.
- [14] J. Koreman, A. C. Morris, D. Wu, S. Jassim, H. Sellahewa, J. Ehlers, G. Chollet, G. Aversano, H. Bredin, S. Garcia-salicetti, L. Allano, B. L. Van, and B. Dorizzi, 'Multimodal biometric authentication on the SecurePhone PDA,' 2006. -
- [15] R. Biddle, S. Chiasson, and P. van Oorschot, 'Graphical passwords: Learning from the first generation,' Technical Report TR-09-09, School of Computer Science, Carleton University , 2009. -
- [30] I. Guyon and A. Elisseeff, 'An introduction to variable and feature selection,' J. Mach. Learn. Res. , vol. 3, pp. 1157-1182, Mar. 2003. -
- [5] Canalys. (2012, Feb.) Smart phones overtake client PCs in 2011. [Online]. Available: http://www.canalys.com/newsroom/smart-phones-overtake-client-pcs-2011 -
- [2] Sophos Naked Security blog, 'Survey says 70% don't password-protect mobiles: download free Mobile Toolkit,' http://nakedsecurity.sophos.com/2011/08/09/ free-sophos-mobile-security-toolkit/ , published Aug 9, 2011. -
- [9] R. Spillane, 'Keyboard apparatus for personal identification,' IBM Technical Disclosure Bulletin , vol. 17, no. 3346, 1975. -
- [1] Norton, 'Norton survey reveals one in three experience cell phone loss, theft,' http://www.symantec.com/about/news/release/article.jsp?prid=20110208\_01 , published Feb 08, 2011. -
Cited by
Items 801–843 of 843
Exploring User Interactions with Commercial Machines via Real-world Application Logs in the Lab
Song, F., Wang, W., Zhou, D., Bryan–Kinns, N., Zhang, J., Chen, Q., & Du, L. (2025) · Journal on Multimodal User Interfaces · doi:10.1007/s12193-025-00456-1 · Cited by 1
“Meanwhile, each operation in Experiment Two was performed 12 times for stable measurements [70].”
LFCA: A Lightweight Framework of Continuous Authentication with High Accuracy
Qin, Y., Tang, Z., Wang, D., & Meng, B. (2024) · International Academic Exchange Conference on Science and Technology Innovation (IAECST) · doi:10.1109/iaecst64597.2024.11118056
“BranRun [14], Antal [15], and Touchalytics [19] record user-device interaction data across various tasks, but they fail to account for volunteer demographics and suffer from sample imbalance, limiting the representation of diverse user characteristics.”
Feature Distribution-Based Touch Biometrics Using CNN and Siamese Networks
Gündüz, G., Kaplan, M., & Temizel, T. T. (2025) · International Workshop on Biometrics and Forensics (IWBF) · doi:10.1109/iwbf63717.2025.11113390
“Unlike methods relying on summary statistics [3], we model complete probability distributions of touch biometrics to capture richer interaction patterns.”
Amulet: Integrating Sensitive Information Access with Identity Authentication on Mobile Devices
Liu, D., Feng, Y., Liu, B., Zhang, T., Chen, H., & Cai, M. (2025) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3749546
“Behavioral features include finger touch dynamics [6, 7], mouth movements during speech [8], typing behaviors [10, 11, 30], and patterns of eye movement [9], etc.”
InfluentialAn Evaluation Framework for Synthetic Touch Gestures – A Deeper Dive
Bachert, M., & Hesenius, M. (2025) · International Conference on Human-Computer Interaction with Mobile Devices and Services Adjunct (MobileHCI Adjunct) · doi:10.1145/3737821.3749563
Touch-based interaction dataset for user behavioral analysis in mobile devices
Garabato, D., Casado, M., Dafonte, C., López-Vizcaíno, M. F., Álvarez, M. A., & Novoa, F. J. (2025) · Data in Brief · doi:10.1016/j.dib.2025.112323
“Some existent datasets, such as Touchalytics [ 8 ], BioIdent Dataset [ 9 ] or BrainRun [ 10 ] offer limited interaction types.”
Adaptive Multi-Model Ensemble Learning for Continuous Behavioral Biometric User Authentication
Tiwari, B., & Kaur, N. (2025) · Asian Conference on Innovation in Technology (ASIANCON) · doi:10.1109/asiancon66527.2025.11280957
An anomaly detection based approach for continuous authentication with smartwatch inertial sensors
Gholami, A., Alaca, F., & Zulkernine, M. (2025) · Computers & Security · doi:10.1016/j.cose.2025.104656 · Cited by 2
Robust Flick-Based Continuous Authentication via CTGAN-Generated Touch Features
Do, T. P., & Yamazaki, Y. (2025) · International Symposium on Computing and Networking Workshops (CANDARW) · doi:10.1109/candarw68385.2025.00054
“Swipe input [13]: Unlike flick input, swipe is not intended for text entry on mobile devices; instead, it is commonly used for tasks such as scrolling through messages or reading documents by performing vertical gestures along the screen, as shown in Figure 1 (b).”
TCTAuth: Triple convolutional transformer-based continuous authentication on smartphones
Li, Z., Liu, G., Chen, Y., Wang, Q., Zhao, X., & Zhang, Y. (2026) · Applied Soft Computing · doi:10.1016/j.asoc.2026.114669
Enhancing Physical Security in Smart Environments with Ambient Intelligence
Nanda, A., Doss, R., Heikamp, F., Kumar, A., Reda, H. T., Anwar, A., Baig, Z. A., Gauravaram, P., Pati, D. P., Kanhere, S. S., et al. (2025) · IEEE International Conference on Trust, Security and Privacy in Computing and Communications (TrustCom) · doi:10.1109/trustcom66490.2025.00078
“Early work in continuous authentication centred on behavioural biometrics, where user identity is inferred from habitual actions such as typing dynamics [4], touchscreen interactions [5], or gait patterns captured through embedded sensors [6], [7].”
Zhao, Y., He, H., Xu, K., Huang, Y., Ren, J., & Su, C. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.6168900
A Dual-Branch Vision Transformer for authenticating users with mouse tracesPreprint
Mazumdar, K., & Sundaram, S. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.6388115
Gbemi, A. A., & Udoh, E. (2025) · International Conference on Computer and Applications (ICCA) · doi:10.1109/icca66035.2025.11430907
“…and the app achieved 99.9% uptime with no crashes during the two-week pilot These results reflect a significant improvement over legacy paper-based systems and even surpass some commercial care platforms that report average upload failure rates of 6–8% and non-GIS-linked swipe verification [38].”
InfluentialGupta, A., Jaiswal, A., & Indra, G. (2025) · International Conference on Advanced Computing and Emerging Technologies (ACET) · doi:10.1109/acet67282.2025.11430240
UserIA: User-Centered Implicit Authentication Leveraging Operant Conditioning
Yang, Y., Huang, X., Riya, F. F., Li, J., & Sun, J. (2026) · IEEE Transactions on Dependable and Secure Computing · doi:10.1109/tdsc.2026.3677093
Shen, Z., Zhao, C. Y., Zou, C., Zhao, X., Zhao, J., & Zou, J. (2026) · IEEE Transactions on Dependable and Secure Computing · doi:10.1109/tdsc.2026.3677366
“In laboratory settings, this fixed-window segmentation strategy may work well because experiments typically require users to perform repetitive tasks, such as browsing photos [19], [22], [25], which usually follow similar behavioral patterns.”
Privacy-Preserving Continuous Authentication of Smartphone Users via Secret Sharing
Hu, M., Wang, D., Han, D., Guo, J., & Tu, B. (2026) · IEEE Transactions on Dependable and Secure Computing · doi:10.1109/tdsc.2026.3680150
“Early research in continuous authentication primarily lever-aged classical machine learning algorithms, such as KNN, Hidden Markov Model (HMM), Naive Bayes (NB), and SVM, to learn user behavioral patterns from touchscreen and motion data [8], [16], [43], [44].”
Modélisation et évaluation de la sécurité des parcours d'authentification
Ndiaye, Y. (2019) · theses.fr (ABES) · doi:10.70675/3f84f61bz80a8z4123z945fz29fc40786ae9
Wan, T., Deng, Y., Wang, C., Pan, J., Guo, Z., Yang, X., & Gao, B. (2026) · IEEE Transactions on Visualization and Computer Graphics · doi:10.1109/tvcg.2026.3680635
“Mobile inputs such as taps, swipes, or dwell times can replace VR gestures, leveraging prior work showing that touch behaviors (pressure, rhythm, path curvature) support reliable user identification [8,10,43].”
Formal Security of Zero Trust Architectures
Poirrier, A. (2024) · theses.fr (ABES) · doi:10.70675/368679edzc429z4365z9ed8z98bd52899a61 · Cited by 7
Typing Patterns Never Lie: Robust Behavioral Biometrics with Hybrid Ensembles
Desta, Z., Dave, R., & Bhavsar, M. (2026) · International Conference on Artificial Intelligence, Computer, Data Sciences and Applications (ACDSA) · doi:10.1109/acdsa67686.2026.11467598
Elouni, A., & Zagrouba, R. (2026) · Lecture notes in networks and systems · doi:10.1007/978-3-032-21147-7_9
LSTM-Driven Behavioral Sequence Modeling for Accidental Payment Detection
Rilwan, M. R. M., S, J. D. P., & Niresh Kumar, S. (2026) · International Electronics & Packaging Technologies Conference: Bridging Skills & Innovation for India's Industry (EPT India) · doi:10.1109/epdmc67535.2026.11496521
“Frank et al. [8] introduced Touchalytics, demonstrating that touchscreen interaction patterns can serve as a reliable behavioral biometric for continuous authentication.”
DgaDetector: Leveraging DNS Analysis for DGA-Based Bot Identification
Tong, V., Mac, H., Pham, H., Quoc, K. N., Dinh, T. T. A., & Tran, D. (2026) · Journal of Network and Systems Management · doi:10.1007/s10922-026-10081-2
“Although the concept of continuous detection has been extensively investigated in previous studies [27 – 30], to the best of our knowledge, DgaDetector is the first attempt to apply this approach to DGA-based botnet detection.”
Beyond Normality: Rethinking Behavioral Biometric Data
Belman, A. K., Chen, F. S., Phoha, V. V., & Mohanty, P. (2025) · Signals and communication technology · doi:10.1007/978-3-031-98036-7_16
Two-level hierarchical touch representation for behavioral biometrics based implicit authentication
Shen, Z., Zhao, C. Y., & Zou, C. (2026) · Applied Soft Computing · doi:10.1016/j.asoc.2026.115625 · Cited by 1
Gesture Based Authentication for the Visually Impaired Using a Smartwatch
Lekkala, S. S., Rahman, K. A., & Mukherjee, A. (2026) · IEEE International Conference on Computing and Machine Intelligence (ICMI) · doi:10.1109/icmi68585.2026.11539872
“Touch-based authentication has shown promise, with [12] demonstrating that touchscreen swipe gestures form unique behavioral signatures, achieving high accuracy in user distinction.”
Erkinuly, M. F., Satybaldiyeva, R., Serikovich, O. G., & Darkhanuly, D. (2026) · Procedia Computer Science · doi:10.1016/j.procs.2026.04.140
Spatio-temporal aware and context contrastive transformer for capacitive touch representation
Shen, Z., Ma, Y., Zou, C., & Zhao, C. Y. (2026) · Applied Soft Computing · doi:10.1016/j.asoc.2026.115631
Swipe Gesture-Based Continuous Mobile Device Authentication: a Systematic Literature Review
Naji, Z., & Bouzidi, D. (2026) · SN Computer Science · doi:10.1007/s42979-026-05090-6
“In [31], the authors used 76 features, selected from 5 papers [ 10, 20, 32 , 34 , 35 ].”
OSD: A Graphical Password Scheme Based on Object Shape Drawing on Smartphones
Berry, C., & Li, W. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-16168-0_9
Aichner, T. (2026) · Springer eBooks · doi:10.1007/978-3-658-52717-4_6
Passing the Turing Test on Screen: A Benchmark for Mobile GUI Agent Humanization
Zhu, J., Yang, L., Shan, R., Zheng, C., Zheng, Z., Liu, W., Yu, Y., Zhang, W., & Lin, J. (2026) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3770855.3817471
“In the mobile domain, detection leverages Behavioral Biometrics , utilizing touch dynamics (e.g., pressure, velocity) for user verification [3, 15, 16, 24, 41, 59].”
Beyond Accuracy: A Cost Aware Evaluation of Mobile Typing-Based Authentication
Desta, Z., Dave, R., & Bhavsar, M. (2026) · International Conference on Electrical, Computer and Energy Technologies (ICECET) · doi:10.1109/icecet65726.2026.11632864
“[9] However, behavioral signals are inherently variable and context-dependent, making robustness and generalizability key challenges.”
Mahramanlıoğlu, T., & Vural, M. S. (2026) · Journal of Computer Virology and Hacking Techniques · doi:10.1007/s11416-026-00667-4
Leveraging Device Motion Events to Track Fraud in Remote Identity Verification
Rosário, L., Boer, S., Madeira, P., Pajot, A., Wozniak, A., & Serrano, D. (2026) · Lecture notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering · doi:10.1007/978-3-032-37274-1_25
Omar, N. S. (2026) · Research Square · doi:10.21203/rs.3.rs-11004507/v1
Jayanthi, M., Browlin Steffy, S., S, M. S. M., Kamesh, M., & R, V. M. (2026) · International Conference Computational Vision and Bio Inspired Computing · doi:10.1109/iccvbic71195.2026.11688892
“Touchalytics showed touchscreen input is a viable continuous biometric [2]; BioPrivacy combines keystroke dynamics and touch gestures for continuous authentication [1]; and feature-level fusion of touch and keystroke reaches ~1% EER with a multilayer perceptron [3].”
Ndalama, F. E., Mutabazi, P., Taenaka, Y., & Kadobayashi, Y. (2026) · IEEE Access · doi:10.1109/access.2026.3734604
“Behavioral biometrics enable unobtrusive continuous authentication using keystroke, mouse, touch, and inertial signals [5], [25], [26], [27], [28], [29], [30], [31].”
Naji, Z., & Bouzidi, D. (2026) · Lecture notes on data engineering and communications technologies · doi:10.1007/978-3-032-18758-1_34
Beyond~Login: Self-Supervised Multimodal Continuous Authentication for Smartphones
Islam, M. T., Alavalapati, G. R., & Odelu, V. (2026) · IEEE Conference on Communications and Network Security · doi:10.1109/cns69863.2026.11720899
“The “Hold On and Swipe” framework combined touchscreen gestures with HMOG sensor data to create a multimodal behavioral authentication system for 51 smartphone users, achieving accuracies of up to 81.74% using Random Forest classifiers [3], [12], [16].”
InfluentialKhedr, W. I. (2026) · Journal of Systems Architecture · doi:10.1016/j.sysarc.2026.104039
Metadata
- Authors
- Frank, M., Biedert, R., Ma, E., Martinovic, I., & Song, D.
- Affiliations
- University of California, Berkeley (US)
- German Research Centre for Artificial Intelligence (DE)
- ORCID
- Frank, Mario
- Venue
- IEEE Transactions on Information Forensics and Security
- Published
- 2013-01
- Volume
- 8
- Issue
- 1
- Pages
- 136-148
- ISSN
- 1556-6013 (ISSN-L)
- 1556-6021
- OpenAlex
- W2151854612
- Semantic Scholar
- 7f42d54971352678ef25ec411fc96cdcc3fdae30
- Cited by
- 844
- References
- 29
- Citations by year
- 2026: 34 · 2025: 32 · 2024: 51 · 2023: 57 · 2022: 62 · 2021: 65 · 2020: 84 · 2019: 67 · 2018: 80 · 2017: 87 · 2016: 89 · 2015: 76 · 2014: 49 · 2013: 9
- FWCI
- 118.14
- Type
- journal article
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- IEEE Transactions on Information Forensics and Security (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Topics
- Physical Sciences › Computer Science › Information Systems › User Authentication and Security Systems
- Physical Sciences › Computer Science › Signal Processing › Biometric Identification and Security
- Physical Sciences › Computer Science › Computer Networks and Communications › Advanced Authentication Protocols Security
- Keywords
- behavioral biometrics, continuous authentication, touch screen input, user identification, multimodal biometric authentication, scrolling behavior
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.