Harvester open publications of NAS Ukraine
Harvester open publications of NAS UkraineNot a member yet
2350 research outputs found
Sort by
Neurosymbolic approach for attack detection in satellite communication systems
Get PDFAbstract: In the context of increasing cyber threats, the pressing task becomes the implementation of new protection systems for satellite communications. The proposed article presents an innovative neurosymbolic method for attack detection that integrates the capabilities of artificial intelligence and neural networks for effective countermeasures against threats in the domain of satellite communication. The foundation of the development is the synthesis of the strengths of symbolic artificial intelligence and deep learning, enabling highly accurate recognition and neutralization of complex attacks. The architecture of the proposed system is thoroughly detailed, including its key components, mechanisms of operation, and implementation process. Analyzing data from satellite and terrestrial networks, the system's effectiveness is evaluated using machine learning methods, demonstrating significant improvements in intrusion detection compared to existing approaches. Special attention is given to the model's ability to adapt to new types of attacks, ensuring its longterm relevance and efficiency. The architecture of the chosen multilayer neural network includes a symbolic layer, designed for analyzing network input data for vulnerabilities or attacks based on a knowledge base. Experiments on datasets of attacks and vulnerabilities such as CTU-13 and STIN allow for the testing and confirmation of the high efficiency of the proposed method. Thus, this research paves the way for improving cybersecurity systems in the field of satellite communication, contributing to the creation of a mor e secure space environment.Prombles in programming 2024; 2-3: 223-23
Neural network application to pseudorandom sequence generation simulation
Get PDFThe article discusses the methodology of using neural networks to simulate pseudorandom sequences, which allows finding the hidden structure and sequence algorithms to reduce the observed processes to deterministic ones. To improve the quality of simulation of sequences of generated numbers, models of recurrent neural networks are used, taking into account their ability to adapt to the generated continuous sequences. The paper proposes a method for using and tuning recurrent neural networks and the influence of selected hyperparameters that determine the internal structure, size of the input sequence, and length of the resulting generated sequence, which ultimately affects the quality and accuracy of the matching neural pseudorandom generator sequences and the real program equivalent. The obtained results show a possibility of using neural methods in the processes of hashing, predicting sequences of statistically random data, cryptographic algorithms and data compression algorithms, or finding the original process of generating the sequence under study.Prombles in programming 2024; 2-3: 280-28
Use of artificial intelligence sn the application for working with musical notes
Get PDFIn the work, existing software solutions and successful IT projects were analyzed and their advantages and disadvantages were identified, which helped determine the requirements for a product that would be competitive and meet the requirements of the modern market. Modeling and designing of the software was carried out, the client-server architecture of the application was described, as well as the interaction of subsystems. The mobile application was developed and tested, and further directions for improvement and development of the application were determined. The application processes a PDF file with a given metronome speed in mp3 and mp4, which gives the user the opportunity to see and listen to the sheet music. The project includes an Android application with a clear and convenient interface, integration with external utilities and libraries. In the work, the processing of files from pdf format to such music and playback files as midi, musicxml, mp3, mp4 is collected in one stream. The process of parsing and playing with full-cycle processing of music files has been improved, by providing the user with all software modules, and the process of processing visual notes and bringing them to easy-to-use files, such as videos that combine notes with sound, has been improved. The work is important because it contributes to the development of digital music processing methods. The introduction of modern technologies for note recognition and visualization of musical elements contributes to technological progress in the field of music development.Prombles in programming 2024; 2-3: 384-39
Integrated open source metadata aggregation environment to support scientific research
Get PDFMetadata for electronic libraries is an integral part of the organization and access to digital resources. In this article, we explore common approaches to e-library integration, focusing on metadata analysis, protocols, and software. The types of metadata considered, such as descriptive, structural, administrative, preservation metadata, rights metadata, and technical metadata, reveal various aspects of digital resources. Various protocols and approaches to integration are considered, in particular the OAI-PMH protocol, which plays a key role in providing access and metadata exchange between repositories. An overview of software for integration using the OAI-PMH protocol is also provided, including a comparative table of characteristics of popular software solutions. This article helps to understand the importance of metadata and the effectiveness of its use to ensure access to digital resources in the modern scientific environment. The processes of extraction, transformation, loading, which can be used in the semantic integration of data through OAI-PMH, are described in detail. Software for the integration of electronic libraries and open access journals was also reviewed and key functions for creating an effective scientific environment were identified.Prombles in programming 2024; 2-3: 408-41
Features of data processing and storage using the virtual file system
Get PDFThe work developed and considered a software system aimed at creating a virtual file system (VFS) based on FUSE, which provides a convenient and flexible interface for interacting with the file structure. The main goal of this project is to speed up file processing by customizing file system operations and creating a tool that allows users to easily manipulate the virtual file structure using a console interface. The development is intended for data processing and storage on operating systems of the Unix family. The functional tasks of the software system are focused on creating a virtual file system (VFS) based on FUSE with the aim of providing users with a convenient interface for interacting with the file structure. The main goal is to provide functionality for creating and deleting objects, obtaining detailed information about file attributes, reading and writing file contents, working with symbolic links, and navigating and interacting with directories. This includes the ability to create new files and directories, delete objects, get detailed information about file attributes, read and edit their contents, and work with symbolic links. The further goal is to create a convenient and functional tool for users, which can be used for conducting experiments, training or solving specific tasks, while providing flexibility and efficiency of interaction with the FUSE-based file system through the console interface.Prombles in programming 2024; 2-3: 334-34
Development of the medicine safety data monitoring system GERMES PV (literature monitoring and internet sources)
Get PDFThe global coronavirus pandemic of recent years has led to an emergency situation in the field of health care, determined the need to develop new vaccines for the treatment of the population under the guidance of the World Health Organization. Assessment of the safety and effectiveness of the use of these vaccines requires a wide range of research, one of the methods of which is literature monitoring of scientific publications. Significant labor costs for the search and analysis of articles make the creation and implementation of computer information and search systems for monitoring literature and Internet sources an urgent issue. Developed for this purpose, the medicine safety data monitoring system Germes PV is intended for the comprehensive application of modern technologies of automated monitoring of periodicals and Internet resources (web portals, websites, electronic libraries, etc.) in order to search for publications on the safety of drugs, availability adverse reactions and lack of effectiveness of medicinal products, feedback and suggestions of interested persons and other information in the field of pharmacovigilance. In this work, a comprehensive analysis of the subject area is carried out and ways of automating the activities of health care institutions and pharmaceutical companies in this area are described.Prombles in programming 2024; 2-3: 190-19
Neuro-symbolic approach for the biological systems and processes research
Get PDFThe current state and prospects for the application of neural network technology and systems built based on neural networks for studying biological systems and processes (in particular, the processes of virus-cell interaction) were considered in the article. In particular, the concept of the neuro-symbolic approach, which combines the use of neural networks and algebraic modelling, was described. The use of an algebraic approach in combination with neural networks allows us to obtain an effective formal proof of biological objects' properties (for example, cells' and viruses' properties) or processes, as well as to search for objects or the necessary values of their parameters that correspond to the specified properties. One of the experiments that we consider is applying the proposed approach to modelling and studying the process of programmed cell death (apoptosis) caused by certain types of viruses and considering the possibility of using the obtained results in medical practice (particularly in the treatment of oncological diseases). The main task of such experiments is to analyze and identify the properties of viruses capable of triggering tumour cell apoptosis and, in fact, to determine the possibility of reaching the final stage of this process under the given parameters of the virus and the cell.Prombles in programming 2024; 2-3: 271-27
The problem of distribution and merging of discrete correspondence flows in individual zones of a hierarchical communication network
Get PDFThe article is devoted to the study of the subproblem of distribution and merging of correspondence flows in separate zones of the backbone network, which arises when solving the general problem of optimizing the hierarchical structure of a multicommodity communication network with discrete flows and parameters. In a multicommodity network, each node can exchange correspondence (products, goods, cargo, messages) with other nodes. Correspondence is characterized by a source node, a drain node and a value, which for transport networks is given by the number of packaged goods in a package of a unified size, and for data transmission networks – by the number of bytes, kilobytes, etc. In the backbone network, all correspondence is transported in vehicles in transport units of a given size (capacity, volume) or transmitted via communication channels. The size of a transport block is measured by the number of units of correspondence that fit into it (for example, 40 packaged goods, 100 gigabytes). All trunk nodes are sorting centers in which correspondence is first sorted by destination addresses (nodes) and then packed as consolidated correspondence into transport blocks. Since the size of individual correspondence is much smaller than the size of the transport block, they can be combined (packed) with correspondence with other destination addresses several times and in different nodes during sorting. There are three levels of hierarchy in the network – backbone, zonal and internal and four types of nodes – trunk nodes of the first, second and third types, forming the backbone and zonal levels of the network and nodes of the fourth type, which are subordinate to each trunk node and form internal levels of the network. Node types differ from each other in functionality. The main task of the study is to develop a mathematical model and algorithms for solving the subproblem of optimizing the distribution and merging (sorting) of correspondence flows at the zonal levels of the network. It is shown that it can be formulated as a linear programming problem with a block structure of constraints and the Danzig-Wolf decomposition method and other methods of integer programming can be used to solve it. To solve the problem on real networks, approximate algorithms based on the construction of the shortest paths are proposed.Prombles in programming 2024; 2-3: 53-6
Reverse synthesis of natural language phrases grounding on their ontological representation using a large language model
Get PDFThe presented article introduces a novel solution that uses a specially developed structured prompt for a large language model (Chat GPT). A series of experiments were carried out on synthesizing natural language phrases based on their ontological representations. These ontological representations were automatically constructed from sentences of scientific and technical texts using previously developed software tools. Such representations contain entities found in the text and typed semantic relationships between them, which can be realised in the phrases of the analysed text. The system of relationships, specified by a set of concepts, is linked with the entity of the related part of the sentence, which in turn can be a simple sentence or part of a complex sentence. The structured prompt for the large language model includes explanations of the semantic relationships between concepts in the context of sentence synthesis from ontological representation, as well as a set of pairs of concepts connected by semantic relationships, which serve as materia l for sentence creation. The synthesised natural language sentences were compared with the originals using the cosine similarity measure across different vectorisation methods. The obtained similarity scores ranged from 0.8193 to 0.9722 according to the xx_ent_wiki_sm model, although stylistic distortions of the generated sentences were observed in some cases. The research presented in this work has practical significance for the development of dialogue information systems that combine the ontological approach with the use of large language models.Prombles in programming 2024; 2-3: 359-368
Global open access resource aggregation services and their requirements for data providers
Get PDFThe paper looks at the modern global aggregators of open access documents BASE, OpenAIRE and CORE. Their quantitative characteristics are analysed, such as the number of collected document descriptions and full texts, the number of data providers, and the availability of an application programming interface for data retrieval. The composition and types of their data providers, such as institutional repositories, open journals, publishing houses, scientific repositories of preprints, thematic digital libraries, as well as systems that are also aggregators, are considered. We also investigate what kind of information about documents these aggregators collect and how it is presented in the user interface, as well as what information is collected about data providers and how it is presented in the user interface. How the aggregator interacts with data providers, what data communication protocols are supported, and how often the collected data is updated. Also, modern aggregators based on collected data corpora, using machine learning methods, bibliometrics, webometrics, altmetrics, semantometrics, provide a range of useful services to researchers. As developers of a number of open access scientific digital libraries, we are already registered as data providers in these systems. Therefore, we are familiar with their requirements in practice. In this article, we have tried to summarise these requirements to some extent, namely: support for standard data transfer protocols, support for standard metadata sets, support for standard character encodings for content, and others.Prombles in programming 2024; 2-3: 449-456