Iryna Kazymyra

The aim of the study is to construct a digital three-dimensional relief model of the Korolivski Beskydy National Nature Park using open geospatial data and QGIS tools, which can serve as a basis for environmental planning and ecological monitoring. Methodology. The methodology of digital terrain modelling for the “Korolivski Beskydy” National Nature Park is based on the built-in algorithms of the QGIS platform, using the Copernicus DEM GLO-30 digital elevation model and the vector layer of the park boundaries. Results. The developed digital relief model of the Korolivski Beskydy NNP reflects the elevation range typical of the mid-mountain Carpathians, with the majority of the area concentrated at medium altitudes. The overall relief variability is confirmed by a moderate spread of elevation values. It was established that the park territory is dominated by moderately steep slopes (5–15°), occupying 54.04% of the area, with a maximum prevalence of southwestern exposures (16.8% of the territory). The calculated hypsometric integral HI = 0.35 indicates an intermediate stage of geomorphological development between the mature and old stages of the erosion cycle. The mean profile curvature indicates a slight predominance of convex landforms, typical of mountainous areas with active denudation. The positive mean plan curvature value indicates a slight dominance of divergent landforms. The conducted morphometric analysis is significant for optimizing park management, enabling hydrological monitoring, forecasting surface runoff on deforested areas, preventing erosion processes, restoring forests, and conserving landscape and biological diversity. Scientific novelty. For the first time, a comprehensive morphometric analysis of the Korolivski Beskydy NNP territory was conducted using the latest Copernicus DEM satellite data, providing a basis for further research on the park’s landscape dynamics. Practical significance. The obtained results will contribute to enhancing the efficiency of environmental protection, recreational, and scientific functions of the Korolivski Beskydy National Nature Park, ensuring the conservation of unique primeval beech-fir forests of the Upper Dnister region and maintaining ecological balance in the headwaters of the Dnister basin.

Monitoring the environmental safety of biosphere reserves is a crucial scientific and applied task, particularly in transboundary regions where natural systems are exposed to complex anthropogenic pressures. The application of Geographic Information Systems (GIS) provides advanced opportunities for integrating heterogeneous spatial data, conducting environmental assessments, and supporting decision-making processes. This study focuses on the development and implementation of an ecological-cartographic model entitled «Settlements and Road Network of Roztochchia», created on the basis of monitoring data collected within the Ukrainian-Polish transboundary Roztochchia biosphere reserve. The proposed interactive mapping framework combines spatial and temporal datasets related to protected areas, settlement patterns, and transportation infrastructure. Such integration enables a more comprehensive understanding of landscape organization and facilitates environmentally balanced land use in border territories. Particular attention is given to the role of the road network as a factor influencing ecosystem fragmentation, accessibility, and anthropogenic transformation of natural complexes. The methodological approach is grounded in a systematic analysis of contemporary scientific research, combined with the application of geoinformation technologies, specialized software tools, and spatial proximity analysis techniques. The ecological-cartographic modeling process was carried out using GIS MapInfo Professional, which allowed for the synthesis and visualization of complex environmental data. The resulting geoinformation resource is designed to improve the efficiency and accuracy of spatial data management related to the nature reserve fund. It also ensures transparency and accessibility of environmental information for the public, while providing analytical support for governmental institutions, stakeholders, and organizations involved in regional planning and sustainable development strategies. In addition, the study substantiates the concept of a road-based protected network aimed at conserving landscape diversity, strengthening ecological connectivity, and enhancing the recreational and tourism potential of the region. The developed geodatabase structure enables the acquisition of reliable and up-to-date information on the condition of protected, economic, and urbanized components of the ecological network. It also supports the planning of forest and hydro-ecological corridors and contributes to the optimization of functional zoning within the Roztochchia biosphere reserve.

The fluorescent parameters of European beech (Fagus sylvatica L.) were investigated in forest communities of the Northern Podillia National Nature Park. The scientific and practical significance of analyzing fluorescent parameters for integrated monitoring of forest ecosystem condition is substantiated. The relevance of monitoring studies of dendroflora lies in establishing a baseline of environmental data to assess the state and dynamics of the park’s ecosystems. Reconnaissance surveys of forest stands were conducted, and fluorescence intensity was measured in trees of different age classes growing under diverse forest ecological conditions. Based on the numerical characteristics of fluorescence intensity dynamics, a tree vitality index and indicator parameters reflecting the influence of exogenous and endogenous factors were calculated. The highest vitality index values were recorded for old-growth trees, attributed to optimal light conditions and site characteristics. In contrast, understory saplings exhibited reduced vitality index values due to decreased light availability, reflecting a decline in the potential activity of the photosynthetic apparatus. Variations in the tree vitality index were observed across different site conditions. The observed decrease in vitality index values is associated with relatively lower soil fertility and water availability, reflecting stress-induced adaptive responses of plant physiological functions to varying environmental conditions and representing a primary cause of photosynthetic inactivation. Based on fluorescence testing and bioinformatic analysis, the physiological state and adaptive potential of European beech within the Northern Podillia National Nature Park were determined. The obtained data on changes in tree fluorescence parameters confirm the photosynthetic apparatus’s sensitivity to environmental factors and substantiate the suitability of European beech as a test object for assessing edaphic and hydrogeological conditions on the territory. It was established that the fluorescence method is promising for monitoring the sanitary condition of forest stands, as it enables the determination of an integral, scientifically informative parameter — the tree vitality index. The practical applicability of modern fluorescence-based express testing methods for vegetation is justified, as these methods provide a correlational synthesis of photobiological mechanisms of photosynthesis and graph-analytical information, which is essential for the development of information and diagnostic systems for integrated ecosystem monitoring, biodiversity conservation, and protection of the floristic gene pool of protected natural areas.

This paper examines the problem of predicting a user’s multidimensional emotional state from textual records under conditions where most existing text-based approaches emphasize either categorical emotion recognition or coarse sentiment polarity, which limits the interpretability of broader affective assessment. The study aims to develop and experimentally validate a model that combines context-sensitive emotion detection based on a BERT classifier fine-tuned on the GoEmotions corpus with a hierarchical aggregation scheme (emotion-to-mood-to-polarity) and a quantitative representation of affect in the Pleasure-Arousal-Dominance space. The empirical basis includes short user-generated texts such as messages, comments, diary fragments, and reflective notes. To ensure compatibility with the English-language emotion corpus, the input text is normalized, translated into English when required, and segmented into paragraphs and sentences. Emotional labels are assigned at the level of contextual sentence windows. These outputs are then aggregated into paragraph-level emotion frequencies, weighted by emotion-specific PAD vectors, and transformed into dominant mood and overall text polarity. The results show that the proposed hierarchy provides a logically consistent transition from local emotional cues to a more stable affective profile. Evaluation on a held-out 10% subset of GoEmotions yielded macro-average Precision, Recall, and F1-score values of 0.518863, 0.485856, and 0.496837, respectively. The highest F1-scores were obtained for gratitude, amusement, love, and remorse, whereas grief, realization, disappointment, and nervousness remained the weakest classes. Scenario-based experiments further showed that the model distinguishes positive, negative, neutral, and mixed texts while preserving affective heterogeneity through PAD-based representation rather than forcing the input into a single categorical state. The contribution of the paper is to substantiate a PAD-oriented and hierarchically aggregated framework for interpretable emotional-state prediction from text, suitable for affective analytics, digital diary analysis, and wellbeing-oriented text assessment, and to demonstrate its potential as an analytical foundation for adaptive, user-centered intelligent systems.

У цьому дослідженні представлено гібридну модель, яка поєднує колаборативну фільтрацію та глибокі нейронні мережі для підвищення ефективності навчальних рекомендацій. Такий підхід дозволяє оптимізувати процес генерації рекомендацій, забезпечуючи високу гнучкість і точність навіть при обмеженій кількості даних. У випадках, коли наявних даних недостатньо для ефективного навчання нейронної мережі, алгоритм автоматично переходить до використання колаборативної фільтрації, що зберігає якість рекомендацій і стабільність роботи моделі. Результати тестування на наборах даних показали, що гібридна модель підвищує точність і адаптивність рекомендаційної системи порівняно з традиційними методами. Поєднання двох методів дозволяє використати їхні переваги та мінімізувати недоліки, що робить модель універсальною в різних контекстах і з різними обсягами даних. Це особливо актуально у сфері освіти, де дані користувачів можуть бути обмеженими або неповними. Впровадження цієї моделі дає можливість створювати більш ефективні освітні платформи, які забезпечують персоналізоване навчання для широкої аудиторії, враховуючи індивідуальні потреби та вподобання користувачів. Важливою особливістю алгоритму є його здатність адаптуватися до змін у навчальному контенті та налаштовувати рекомендації відповідно до потреб кожного користувача. Це сприяє підвищенню задоволеності користувачів і покращує загальну якість освітніх послуг. Ефективна робота з обмеженими даними підвищує точність і адаптивність системи рекомендацій. Усі ці властивості роблять запропонований алгоритм перспективним інструментом для розвитку персоналізованих освітніх платформ, які краще відповідають індивідуальним потребам. Це сприяє ефективнішому навчанню та кращому засвоєнню матеріалу, що є важливим для розвитку сучасного інформаційного суспільства

Функціонування складних систем часто супроводжується виникненням збоїв, спричинених аномальною чи критичною поведінкою окремих компонентів. Актуальним завданням є моніторинг змін стану систем з метою вчасного виявлення нештатних ситуацій і запобігання збоям. Сучасні методи моніторингу часто не враховують динамічну природу складних обчислювальних систем і зазвичай ігнорують користувачів як активних учасників. Використання штучного інтелекту допомагає покращити процес моніторингу та зробити його ефективнішим. У даній роботі представлено підхід, що дозволяє фіксувати динаміку змін властивостей складних систем та виявляти аномалії, які можуть свідчити про наближення системних збоїв.

The paper highlights the increasing importance of machine learning (ML) in mobile applications, with mobile devices becoming ubiquitous due to their accessibility and functionality. Various ML models, including Convolutional Neural Networks (CNNs) and Vision Transformers (ViTs), are explored for their applications in real-time classification on mobile devices. The paper identifies key challenges in deploying these models, such as limited computational resources, battery consumption, and the need for real-time performance. Central to the research is the comparison of MobileNetV2, a lightweight CNN designed for mobile applications, and Vision Transformers (ViTs), which have shown success in image recognition tasks. MobileNetV2, with its depthwise separable convolutions and residual connections, is optimized for resource efficiency, while ViTs apply self-attention mechanisms to achieve competitive performance in image classification. The study evaluates the performance of both models before and after applying optimization techniques like quantization and graph optimization. It was discovered that quantization is one of the most effective optimization strategies for mobile environments, reducing model size by up to 74 % and improving inference speed by 44 % in ViTs. Additionally, graph optimization techniques, such as operator fusion, pruning, and node reordering, are examined for their role in reducing computational complexity and improving performance on resource-constrained devices. Experimental results on different datasets, including MNIST and the ASL Alphabet dataset, demonstrate the significant performance improvements achieved through optimization. The study shows that post-training quantization and graph optimization can reduce model size, inference time, and CPU usage, making ML models more suitable for mobile applications. The experiments were conducted on a Xiaomi Redmi Note 8 Pro device, showcasing the practical benefits of these optimizations in real-world mobile deployments. The research concludes that optimization techniques like quantization and graph optimization are essential for deploying ML models on mobile devices, where resource constraints and real-time performance are critical. It also provides valuable insights into how ML architectures can be optimized for mobile environments, contributing to the advancement of efficient AI-driven mobile applications.

The work solves the urgent task of improving professional software for statistical text analysis in accordance with the requirements of specialists. Techniques and perspectives of statistical research in linguistics were analyzed and information technology (IT) was developed to determine the statistical profile of Ukrainian-language texts. Comprehensive work on the modeling of the software system was carried out, presented in the form of diagrams that holistically reflect the functioning and purpose of the developed product. The components of information technology were developed and the software system's structure based on the modular principle was proposed. Mathematical models and algorithms have been developed, that make it possible to determine the main characteristics (statistical profile) of the studied Ukrainian texts. They were implemented in software designed using Python. The results of the study of Ukrainian-language texts and their statistical profiles are given, it is demonstrated that advanced information technologies provide processing of Ukrainian-language texts with a high level of automation. The obtained results can be considered as a contribution to the development of scientific research in applied linguistics, thanks to which conditions are created for studying authors' texts of various styles and the effective use of professional skills and knowledge by a wide range of users.

An approach to the implementation of a neural network for real-time cryptographic data protection with symmetric keys oriented on embedded systems is presented. This approach is valuable, especially for onboard communication systems in unmanned aerial vehicles (UAV), because of its suitability for hardware implementation. In this study, we evaluate the possibility of building such a system in hardware implementation at FPGA. Onboard implementation-oriented information technology of real-time neuro-like cryptographic data protection with symmetric keys (masking codes, neural network architecture, and matrix of weighting coefficients) has been developed. Due to the pre-calculation of matrices of weighting coefficients and tables of macro-partial products and the use of tabular-algorithmic implementation of neuro-like elements and dynamic change of keys, it provides increased cryptographic stability and hardware–software implementation on FPGA. The table-algorithmic method of calculating the scalar product has been improved. By bringing the weighting coefficients to the greatest common order, pre-computing the tables of macro-partial products, and using operations of memory read, fixed-point addition, and shift operations instead of floating-point multiplication and addition operations, it provides a reduction in hardware costs for its implementation and calculation time as well. Using a processor core supplemented with specialized hardware modules for calculating the scalar product, a system of neural network cryptographic data protection in real-time has been developed, which, due to the combination of universal and specialized approaches, software, and hardware, ensures the effective implementation of neuro-like algorithms for cryptographic encryption and decryption of data in real-time. The specialized hardware for neural network cryptographic data encryption was developed using VHDL for equipment programming in the Quartus II development environment ver. 13.1 and the appropriate libraries and implemented on the basis of the FPGA EP3C16F484C6 Cyclone III family, and it requires 3053 logic elements and 745 registers. The execution time of exclusively software realization of NN cryptographic data encryption procedure using a NanoPi Duo microcomputer based on the Allwinner Cortex-A7 H2+ SoC was about 20 ms. The hardware–software implementation of the encryption, taking into account the pre-calculations and settings, requires about 1 msec, including hardware encryption on the FPGA of four 2-bit inputs, which is performed in 160 nanoseconds.