Year: 2026
Type: dissertation
Abstract: The paper addresses the problem of ensuring reliable protection of information systems against unauthorized access and session hijacking. An improved method for continuous user authentication based on the comprehensive analysis of behavioral biometric characteristics specifically keystroke dynamics and mouse movement patterns is proposed. The method involves using machine learning algorithms for the simultaneous processing of both data types and the generation of a unique behavioral profile in the background. The implementation of the proposed approach will improve the accuracy of legitimate user identification, reduce the false acceptance rate, and provide imperceptible verification throughout the entire session.
Authors: Білоус, В. М., Дзюбенко, К. Р.
Language: Ukrainian
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7213720652
