Users identification by keyboard handwriting using neural networks
Article Sidebar
Main Article Content
Abstract
In our time of digital technologies, one of the leading tasks in cybersecurity is reliable user identification. In recent years, biometric methods that allow for clear user identification have been rapidly gaining popularity. One of the most common cybercrimes is password theft and logging in with these data to various services: banking, insurance, education, tax. There, cybercriminals, gaining rights and access to user information, cause citizens financial, reputational and moral losses. Private and public institutions can suffer even greater losses due to the leakage of personal data, private information about property and financial status. Every year, the power and number of cyberattacks using stolen passwords only increases. Therefore, it is necessary to develop a set of methods and measures to counter such cyberattacks. One of the biometric methods for user identification is the behavioral biometric method using keyboard handwriting, which is an individual feature of each person. To obtain the characteristics of keyboard handwriting, no additional sensors are required - a regular keyboard and some software are enough. The article considers the main characteristics of keyboard handwriting, a neural network is used to recognize the author's and non- author's password entry. In general, it was possible to build a fairly simple and effective neural network for identifying users by keyboard handwriting. The use of this method can significantly enhance the reliability of user identification and cybersecurity of information telecommunication systems.
Article Details
References
2. Ivanov A.I. Biometric identification of a person based on the dynamics of subconscious movements. ‒ Penza: Publishing house Penza State University, 2000. ‒ 188 p.
3. Chalaya L.E. User identification model based on keyboard handwriting. «Artificial Intelligence», No. 4. 2004, Р. 811-817.
4. Bryan W.L., & Harter N. (1897). Studies in the physiology and psychology of the telegraphic language. Psychological Review, 4(1), 27-53. ‒ https://doi.org/10.1037/h0073806
5. L.H. Shaffer Reading and Typing ‒https://www.researchgate.net/publication/233266615_Reading_and_Typing
6. Saket Mahesh wary, Soumyajit Ganguly, Vikram Pudi. DeepSecure: A FastandSimpleNeuralNetworkbasedapproachfor User Authentication and Identification via Key stroke Dynamics. https://www.researchgate.net/publication/322952671_Deep_Secure_A_Fast_and_Simple_Neural_Network_based_approach_for_User_Authentication_and_Identification_via_Keystroke_D ynamics.
7. Bryuhomitsky, Yu.A. Method of training neural network biometric systems taking in to account copying of regions/Ya.A. Bryuhomitsky, M.M. Kazarin. – Promising in for mation technologies and intelligent systems (Electronic journal). ‒ 2003. ‒ No. 3 (15). ‒ Р. 17-23.
8. Litvinchuk I.S, Korchomny R.O., Borisov I.V., Korshun N.V. Development of recommendations for minimizing the risks of hacking credentials based on the analysis of the most common hacking methods. Cybersecurity: education, science, technology, No. 4-12, 2021, Р. 163-171.
9. Back propagation and stochastic gradientdes cent method / Amari S. // Neurocomputing – 1993. – № 5. – p. 185-96.
10. Stochastic gradient learning in neural networks / Bottou L. // Proceeding sof Neuro-Nımes –1991. – № 91 – 12 p.
11. Nielsen M. A. Neural Networks and Deep Learning. – Determination Press, 2015.
12. TheDeltaRule [Electronicresource] / Russell I. – University of Hartford, 2012 – Mode of access: https://web.archive.org/web/20160304032228/http:/uha-vax.hartford.edu/compsci/neural-networks-delta-rule.html. – Dateofaccess: 12.06.2024.
13. Danylyuk I.I., Karpinets V.V., Priymak A.V., Yaremchuk Yu.E. Kostyuchenko O.I. Method of user identification by key board hand writing based on neural networks. pp. Registration, storageanddataprocessing, 2018, Vol. 20 No. 2, Р. 68-76.