Search academic texts

Registration Number

0225U003454, (0124U003899) , R & D reports

Title

Models and methods of processing information processes in complex networks to improve the quality of service and information security

popup.stage_title

Моделі та методи обробки інформаційних процесів у складних мережах для підвищення якості надання послуг та інформаційної безпеки

Head

Meleshko Yelyzaveta V., Доктор технічних наук

Registration Date

30-06-2025

Organization

Meleshko Yelyzaveta V.

popup.description1

Development of models and methods of processing information processes in complex networks, represented by computer networks, social networks and recommendation systems, for to improve the quality of providing services to their users and to increase the level of cybersecurity in them

popup.description2

Object of the research: processing of information processes in complex networks. Purpose of the research: development of models and methods for processing information processes in complex networks, represented by computer, social, and recommendation networks, aimed at improving service quality and information security within them. Scientific results: For the first time, a method of simulation modeling of complex networks has been developed, which, unlike existing approaches, generates a hybrid network graph from multiple subgraphs and simulates the behavior not only of regular users, but also of bots and information attacks. This enabled effective testing of computer network methods. For the first time, a mathematical model for detecting anomalous connections in computer systems was developed using dynamic chaos theory, which, unlike existing ones, employs a geometric approach based on changes in mutual orthogonality between components, reducing anomaly detection time without losing accuracy. For the first time, a mathematical model for assessing the current state of a computer system was developed using dynamic chaos theory, which, unlike existing ones, allows representation of temporal changes in component interconnections. For the first time, a method was proposed for determining the probability dynamics of states of information-recommendation system using Markov and semi-Markov processes for optimizing system maintenance costs under internal and external destabilizing factors. For the first time, a method for detecting information attacks in recommendation systems was developed through analysis of object rating trends, reducing security monitoring costs. For the first time, a method for detecting botnets in recommendation systems was developed using graph clustering, enabling neutralization of information attacks.

Product Description

popup.authors

Meleshko Yelyzaveta V.

popup.nrat_date

2025-06-30