NEURAL NETWORK METHOD OF USER BEHAVIOR ANALYSIS BASED ON KEYSTROKE DYNAMICS
Abstract
The reason for most of the information leaks is the compromise of account data. In this regard,
the introduction of additional means of identification and authentication is relevant. To increase
efficiency, these systems are developed using machine learning. The use of neural networks
is currently the most promising approach to improving the security of systems due to their speed
and accuracy. This paper discusses the use of keystroke dynamics to identify authorized users.
Artificial neural networks are used to analyze the dynamics of pressing. In this paper, such characteristics
as the time of pressing a key, the time between keystrokes, the time between releasing
the first key and pressing the second were analyzed. Both convolutional and recurrent neural networks
were used. The primary processing of input data was carried out by a sliding window that
formed data blocks of a certain size. For further processing of already structured data, a onedimensional
convolutional neural network was chosen, since it is well suited for processing data
presented in the form of a sequence. A recurrent neural network, namely the LSTM architecture,
was used to process time dependencies, since it processes variable-length sequences best and is
less susceptible to gradient decay and explosion than others. For experimental verification of the
effectiveness of this technique, the following architectures were implemented: 2xLSTM, 1D SNC +
LSTM, 1D SNC + 2xLSTM. Based on the results of model training, it was revealed that the system
based on the 1D SNC + 2xLSTM architecture with a sliding window size of 50 has the highest
accuracy. The validation accuracy of this architecture was 98.29%. ROC curves were constructed,
which confirmed the effectiveness of this architecture. The F-measure was calculated, which
showed that the highest performance of binary classification is achieved when using the 1D SNC
+ 2xLSTM architecture with a sliding window size of 50 and equal to 99.39%.
References
neyronnykh setey [Deep learning. Dive into the world of neural networks], 2018, 481 p.
2. When Simple Statistical Algorithms Outperform Deep Learning: A Case of Keystroke Dynamics
// Ahmed Anu Wahab, Daqing Hou. Available at: https://www.scitepress.org/Papers/
2023/116841/116841.pdf (accessed 1 June 2023).
3. Chastikova V.A., Vasil'ev E.D., Babich D.V. Neyrosetevaya metodika identifikatsii lits v
videopotoke v usloviyakh ogranichennosti dannykh [Neural network technique for identifying
faces in a video stream in conditions of limited data], Vestnik Adygeyskogo gosudarstvennogo
universiteta. Seriya 4: Estestvenno-matematicheskie i tekhnicheskie nauki [Bulletin of the
Adygea State University. Series 4: Natural, mathematical and technical sciences], 2019, No. 3
(246), pp. 85-90.
4. Chastikova V.A., Titova A.A., Voylova D.O. Analiticheskiy obzor metodov identifikatsii
lichnosti na osnove biometricheskikh kharakteristik [Analytical review of personal identification
methods based on biometric characteristics], Vestnik Adygeyskogo gosudarstvennogo
universiteta. Seriya 4: Estestvenno-matematicheskie i tekhnicheskie nauki [Bulletin of the
Adygea State University. Series 4: Natural, mathematical and technical sciences], 2022, No. 1
(296), pp. 92-112.
5. Dannye [Data]. Available at: https://www.tadviser.ru/index.php/Stat'ya:Dannye (accessed
20 April 2023).
6. Understanding LSTM Networks. Available at: http://colah.github.io/posts/2015-08-
Understanding-LSTMs/ (accessed 1 June 2023).
7. Chastikova V.A., Zherlitsyn S.A., Volya Ya.I., Sotnikov V.V. Neyrosetevaya tekhnologiya
obnaruzheniya anomal'nogo setevogo trafika [Neural network technology for detecting anomalous
network traffic], Prikaspiyskiy zhurnal: upravlenie i vysokie tekhnologii [Caspian Journal:
management and high technologies], 2020, No. 1 (49), pp. 20-32.
8. Basic CNN Architecture: Explaining 5 Layers of Convolutional Neural Network Available at:
https://www.upgrad.com/blog/basic-cnn-architecture/ (accessed 1 June 2023).
9. Gelig A.Kh., Matveev A.S. Vvedenie v matematicheskuyu teoriyu obuchaemykh
raspoznayushchikh sistem i neyronnykh setey: ucheb. posobie: monografiya [Introduction to
the mathematical theory of learning recognition systems and neural networks: textbook. manual:
monograph]. Moscow: Izd-vo SPbGU, 2014, 224 p.
10. Kruglov V.V., Borisov V.V. Iskusstvennye neyronnye seti. Teoriya i praktika: monografiya
[Artificial neural networks. Theory and practice: monograph]. 2nd ed. Moscow: Goryachaya
liniya – Telekom, 2002, 382 p.
11. Fransua Sholle. Glubokoe obuchenie na Python [Deep learning in Python]. St. Petersburg:
Izd-vo «P ter», 2020, 576 p.
12. Gudfellou Yan, Bendzhio Ioshua. Glubokoe obuchenie [Deep learning]. Moscow: Izd-vo
«DMK Press», 2017, 654 p.
13. Vorona V.A. Biometricheskaya identifikatsiya lichnosti [Biometric personal identification].
Moscow: Izd-vo «Goryachaya L n ya – Teleko », 2021, 228 p.
14. Khaykin S. Neyronnye seti. Polnyy kurs [Neural networks. Full course]. 2nd ed. Moscow:
Vil'yams, 2018, 1104 p.
15. Ryshard T. Elementarnoe vvedenie v tekhnologiyu neyronnykh setey s primerami program
[An elementary introduction to neural network technology with example programs], Moscow,
2011, 408 p.
16. Pal S., Dzhulli A. Biblioteka Keras – instrument glubokogo obucheniya [The Keras library is a
deep learning tool]. Moscow: DMK Press, 2018, 294 p.
17. Vasil'ev A.N., Tarkhov D.A. Printsipy i tekhnika neyrosetevogo modelirovaniya [Principles and
techniques of neural network modeling]. Moscow: Vysshaya shkola, 2014, 218 p.
18. Chastikova V.A., Sotnikov V.V. Method of analyzing computer traffic based on recurrent neural
networks, Journal of Physics: Conference Series. International Conference "High-Tech and
Innovations in Research and Manufacturing," HIRM 2019, 2019, pp. 012133.
19. Chastikova V.A., Zherlitsyn S.A. Neyrosetevaya metodika identifikatsii lichnosti po risunku
ven ladoni [Neural network technique for identifying a person by the pattern of the veins of the
palm], Izvestiya YuFU. Tekhnicheskie nauki [Izvestiya SFedU. Engineering Sciences], 2022,
No. 4 (228), pp. 192-200.
20. Chastikova V.A., Zherlitsyn S.A., Voylova D.O. Neyrosetevaya sistema biometricheskoy
identifikatsii lichnosti po golosu [Neural network system for biometric identification of a person
by voice], Vestnik Adygeyskogo gosudarstvennogo universiteta. Seriya 4: Estestvennomatematicheskie
i tekhnicheskie nauki [Bulletin of the Adygea State University. Series 4:
Natural, mathematical and technical sciences], 2023, No. 1 (316), pp. 70-79.
21. Artyushina L.A., Troitskaya E.A. Nekotorye podkhody k otsenke informativnosti parametrov
identifikatsii pol'zovateley po klaviaturnomu pocherku na osnove povedencheskoy biometrii
[Some approaches to assessing the information content of user identification parameters based
on keyboard handwriting based on behavioral biometrics], Vestnik Yuzhno-Ural'skogo
gosudarstvennogo universiteta. Seriya: Komp'yuternye tekhnologii, upravlenie,
radioelektronika [Bulletin of the South Ural State University. Series: Computer technologies,
control, radio electronics], 2022, No. 22, pp. 30-38.
22. Sharipov R.R., Sitnikov A.N. Problemy pri razrabotke sistem raspoznavaniya pol'zovateley po
klaviaturnomu pocherku [Problems in developing user recognition systems based on keyboard
handwriting], Vestnik tekhnologicheskogo universiteta [Bulletin of the Technological University],
2019, Vol. 22, No. 10, pp. 143-147.
23. Eremenko Yu.I., Olyunina Yu.S. Ob opredelenii naibolee znachimykh parametrov
klaviaturnogo pocherka s pomoshch'yu regressionnogo analiza [On determining the most significant
parameters of keyboard handwriting using regression analysis], Sistemy upravleniya i
informatsionnye tekhnologii [Control systems and information technologies], 2018, No. 2 (72),
pp. 28-31.
