DSpace Зібрання:

Еволюція текстових моделей: рекурентні нейронні мережі в контексті інформаційної безпеки

Thu, 01 Jan 2026 00:00:00 GMT

Назва: Еволюція текстових моделей: рекурентні нейронні мережі в контексті інформаційної безпеки Автори: Почанський О. Короткий огляд (реферат): У статті проведено комплексний аналіз методів автоматичної генерації тексту за допомогою рекурентних нейронних мереж (RNN) та моделей довгої короткочасної пам’яті (LSTM), а також досліджено сучасні інтелектуальні й інфраструктурні технології протидії спаму. Розглянуто структурні особливості LSTM-мереж та їхню здатність моделювати довгострокові контекстні залежності при створенні текстів. Систематизовано актуальні категорії спаму та здійснено порівняльний аналіз технічних засобів захисту електронної пошти, зокрема криптографічних протоколів автентифікації відправників SPF, DKIM, DMARC та методу сірого списку (greylisting). Обґрунтовано математичний підхід до застосування наївного байєсівського класифікатора для контентної фільтрації повідомлень. Результати роботи мають практичне значення для проектування комбінованих систем кібербезпеки та захисту інформаційних комунікацій від автоматизованого спаму і фішингу.

Development of a system of indices for monitoring and assessing the sustainability of underground utilities

Thu, 01 Jan 2026 00:00:00 GMT

Назва: Development of a system of indices for monitoring and assessing the sustainability of underground utilities Автори: Liubynskyi P. Короткий огляд (реферат): The paper proposes a comprehensive system of indices to monitor and assess the operational sustainability of urban underground utility networks. Modern municipal engineering infrastructures face critical issues such as accelerating infrastructure aging, highly aggressive operating environments, and stringent financial constraints. To systematically address these challenges, this study identifies seven distinct categories of performance indices: availability, funding sources, effectiveness of rehabilitation work, accident mitigation, environmental safety, efficient use of funds, and efficiency of monitoring implementation. Each index is formalized mathematically to measure a specific operational or socio-economic dimension of utility networks. The developed methodology establishes strict regulation for calculating these indicators, determines their variations across stable intervals, and uncovers synergistic correlations between proactive organizational-technological monitoring and long-term network durability. The practical significance lies in providing municipal utility managers with a reliable decision-making framework to optimize capital investments, mitigate environmental hazards, and sustain public service provision stability.

Prerequisites for developing a business process automation system for an IT enterprise

Thu, 01 Jan 2026 00:00:00 GMT

Назва: Prerequisites for developing a business process automation system for an IT enterprise Автори: Pochanskyi O. Короткий огляд (реферат): The article examines the prerequisites for creating and implementing business process automation systems at information technology (IT) enterprises. The evolution of IT infrastructure management approaches from a resource-based model to a service-oriented concept (ITSM) based on the ITIL library of best practices is analyzed. The architectural features of building modern information systems, particularly the client-server architecture, as well as tools for modeling business processes and databases (ERwin, BPwin, PowerDesigner) are investigated. A comparative analysis of current ServiceDesk/HelpDesk software solutions available on the market (BPM'online service, ITSM 365, ServiceNow, ITSM InfraManager) is conducted, highlighting their advantages, disadvantages, and pricing characteristics. The necessity of developing affordable specialized automation systems for small IT businesses is substantiated.

Hybrid Rule-Based / Machine-Learning Autoscaler for Kubernetes: Reducing Resource Over-Provisioning on a Real Cluster

Thu, 01 Jan 2026 00:00:00 GMT

Назва: Hybrid Rule-Based / Machine-Learning Autoscaler for Kubernetes: Reducing Resource Over-Provisioning on a Real Cluster Автори: Murzha D. Короткий огляд (реферат): Resource allocation for microservice applications in Kubernetes is still handled, in most deployments, by reactive mechanisms. The Horizontal Pod Autoscaler (HPA) and extensions such as KEDA change the replica count only after a monitored metric has already crossed a fixed threshold, which in practice means over-provisioning under noisy load and a sluggish response when demand genuinely spikes. This paper presents HMRO (Hybrid Microservices Resource Optimizer), an autoscaler that pairs a deterministic rule-based engine with an ensemble of machine-learning load predictors whose influence on each decision is not fixed but adjusted continuously according to how accurate the predictors have recently been. HMRO was evaluated on a real Kubernetes cluster (Minikube) against the standard HPA across memory and combined CPU+memory workloads, each with three scenarios and ten iterations. HMRO reduced the average replica count (a direct proxy for over-provisioning) by 36–42% for memory (all p < 0.05) and 20–28% for combined workloads (significant in two of three scenarios), while triggering a comparable number of scaling actions – that is, without losing responsiveness. A comparable reduction was observed for CPU-driven workloads in an earlier evaluation on a prior prototype version. An ablation study isolates the source of the saving: it comes primarily from the asymmetric rule engine, whereas the ML component adds proactivity rather than resource reduction.