Scientific Journal of Astana IT University
Not a member yet
250 research outputs found
Sort by
COMPARATIVE EFFECTIVENESS OF RULE-BASED AND MACHINE LEARNING METHODS IN SENTIMENT ANALYSIS OF KAZAKH LANGUAGE TEXTS
Get PDFSentiment analysis is increasingly pivotal in natural language processing (NLP), crucial for deciphering public opinions across diverse sectors. This research conducts a comparative examination of rule-based and machine learning (ML) methods in sentiment analysis, specifically targeting the Kazakh language. Given the Kazakh language's limited exposure in computational linguistics, the study meticulously evaluates datasets from news articles, literature, and Amazon product reviews, aiming to compare the efficiency, adaptability, and overall performance of these distinct approaches.
Employing a detailed set of evaluation metrics such as accuracy, precision, recall, and computational efficiency, the study provides a comprehensive analysis of the strengths and limitations of rule-based techniques versus ML models like Logistic Regression, Multinomial Naive Bayes, Decision Trees, Random Forest, and XGBoost. The findings suggest rule-based methods excel in identifying nuanced emotional expressions within literary texts, while ML models demonstrate superior adaptability and robustness, particularly effective in handling the linguistic variations found in news and reviews.
Despite the strengths identified, the study also reveals significant limitations of the rule-based approach, especially in broader contexts beyond literary analysis. This highlights an imperative for future research to integrate sentiment dictionaries or domain-specific lexicons that cater to a wider array of linguistic styles, potentially enhancing sentiment analysis tools' applicability in Kazakh and similar less-studied languages.
This investigation contributes significantly to the sentiment analysis discourse, offering invaluable insights for both researchers and practitioners by elucidating the complexities of applying NLP technologies across diverse linguistic landscapes, thus advancing the understanding and methodologies of sentiment analysis in the Kazakh language context
SCIENTIFIC ASPECTS OF MODERN APPROACHES TO MACHINE TRANSLATION FOR SIGN LANGUAGE
Get PDFScientific research in the field of automated sign language translation represents a crucial stage in the development of technologies supporting the hearing-impaired and deaf communities. This article presents a comprehensive approach to addressing semantic and technical challenges associated with the uniqueness of sign language. The research goal is to create an innovative system that combines semantic analysis, sign synthesis, and facial mimicry for the most accurate conveyance of emotional context. The study focuses on the unique features of the Kazakh language and cultural contexts that influence sign communication. The research centers on the development of a semantic system capable of adequately interpreting metaphors, idioms, and classifier predicates of sign language. The three-dimensional nature of signs is analyzed, and a solution to the formal description problem is proposed. The article introduces a database, analysis algorithm, and a prototype 3D avatar capable of translating textual data into sign language. Special attention is given to the processing of idioms and variability in expressing emotions in sign language. Utilizing machine learning principles and computational linguistics algorithms, the authors present an integrated approach to sign language translation, considering linguistic, cultural, and emotional aspects. The proposed algorithms and formulas facilitate effective interaction between textual information and sign expression. The research results not only provide an overview of current challenges in automated sign language translation but also offer practical approaches to addressing them. The developed approach could be a key step towards creating more efficient communication systems for the hearing-impaired and deaf. Which in the future may solve numerous issues with Kazakh sign language
MODELING OF BLOOD FLOW IN LAMINAR MODE
Get PDFThis article presents a detailed analytical evaluation and comprehensive description of a mathematical model designed to simulate blood flow within the human cardiovascular system. The primary objective of this research is to develop a computational model capable of accurately simulating blood flow dynamics and to assess the variations in results using different numerical methods for solving the Navier-Stokes equations, which govern fluid motion. To achieve this, the study begins with an in-depth examination of the anatomy of the cardiovascular system, including various cardiovascular diseases such as stenosis and atherosclerosis, which significantly affect blood flow. The model incorporates important characteristics of blood, treating it as a viscous fluid under laminar flow conditions. Using the Navier-Stokes equations, it was developed a Python-based model to simulate these flow conditions and solve for different flow variables, such as velocity and pressure fields, under both normal and pathological conditions. The computational model was developed using two numerical methods: the Euler method and the Alternating Direction Implicit (ADI) method, which were compared in terms of their computational efficiency and accuracy. The simulations generated insights into how plaque buildup (stenosis) affects blood flow by altering wall shear stress and velocity profiles. This model, while built on foundational fluid dynamics principles, serves as an essential step towards creating a virtual reality (VR) surgical simulator for cardiovascular procedures. This simulator aims to assist surgeons in visualizing and planning surgical interventions by providing an interactive and realistic environment for studying blood flow and related complications
IMPLEMENTATION OF NATURAL EXPERIMENTS IN PHYSICS USING COMPUTER VISION
Get PDFLaboratory experiments in physics are a fundamental basis for studying physical phenomena occurring in nature and a methodological tool that provides visibility of the learning process and conducting experiments is important for the formation of students’ scientific worldview, deep understanding of physical laws and increasing interest in the study of physics. Existing in universities and schools, in addition to traditional ones, modern tools, technologies and approaches, such as virtual reality, augmented reality, computer modelling, online laboratories, virtual laboratory and others, are additional tools for improving the quality of the learning process and teaching techniques, which do not replace full-scale experiments, but only supplement them. In our opinion, for better learning, laboratory installations in physics are needed, with the help of which students can carry out real-life experiments and can broadcast them using innovative computer technologies for distance learning. To implement this task, we reviewed and analysed existing laboratory installations, identified their advantages and disadvantages, and then designed and developed alternative digital experimental set-ups for studying physics phenomena in laboratory conditions of educational institutions based on computer vision technology and presented the results of the study in this article. In carrying out the research tasks, effective methods of conducting scientific research were used, such as theoretical substantiation of the issue, experimental testing of the developed hardware and software systems and computer final processing of experimental data. In summary, the research described in the paper presents an innovative mechanism for integrating object tracking based on computer vision to improve the quality of measurements and new ways of conducting physics experiments. The mechanical laboratory complexes we have developed consist of hardware and software parts. The software part consists of server and client parts. The hardware consists of the main part - the scene, where the physical process takes place, i.e. where a physical object is located, such as a mathematical pendulum, an inclined plane, etc., with the help of which many physical phenomena and processes in mechanics can be demonstrated, and an additional part where a microcomputer and a camera are located. The operating principle of the laboratory installation is based on the use of computer vision technology, i.e. a system for monitoring the ongoing physical process, consisting of a digital camera for image processing, object identification and data export, and a microcomputer for processing experimental data. The use of the experimental installations in the process of teaching physics is a new model of teaching with a promising future in secondary and higher education, and the installations themselves will become tools for offline and online learning, due to the use of computer vision technology, revealing new opportunities and approaches to teaching
COMPARATIVE ANALYSIS OF MULTILINGUAL QA MODELS AND THEIR ADAPTATION TO THE KAZAKH LANGUAGE
Get PDFThis paper presents a comparative analysis of large pretrained multilingual models for question-answering (QA) systems, with a specific focus on their adaptation to the Kazakh language. The study evaluates models including mBERT, XLM-R, mT5, AYA, and GPT, which were tested on QA tasks using the Kazakh sKQuAD dataset. To enhance model performance, fine-tuning strategies such as adapter modules, data augmentation techniques (back-translation, paraphrasing), and hyperparameter optimization were applied. Specific adjustments to learning rates, batch sizes, and training epochs were made to boost accuracy and stability. Among the models tested, mT5 achieved the highest F1 score of 75.72%, showcasing robust generalization across diverse QA tasks. GPT-4-turbo closely followed with an F1 score of 73.33%, effectively managing complex Kazakh QA scenarios. In contrast, native Kazakh models like Kaz-RoBERTa showed improvements through fine-tuning but continued to lag behind larger multilingual models, underlining the need for additional Kazakh-specific training data and further architectural enhancements. Kazakh’s agglutinative morphology and the scarcity of high-quality training data present significant challenges for model adaptation. Adapter modules helped mitigate computational costs, allowing efficient fine-tuning in resource-constrained environments without significant performance loss. Data augmentation techniques, such as back-translation and paraphrasing, were instrumental in enriching the dataset, thereby enhancing model adaptability and robustness. This study underscores the importance of advanced fine-tuning and data augmentation strategies for QA systems tailored to low-resource languages like Kazakh. By addressing these challenges, this research aims to make AI technologies more inclusive and accessible, offering practical insights for improving natural language processing (NLP) capabilities in underrepresented languages. Ultimately, these findings contribute to bridging the gap between high-resource and low-resource language models, fostering a more equitable distribution of AI solutions across diverse linguistic contexts
ANALYSIS AND ASSESSMENT OF AIR QUALITY IN ASTANA: COMPARISON OF POLLUTANT LEVELS AND THEIR IMPACT ON HEALTH
Get PDFThis study presents an in-depth analysis of air quality in Astana, Kazakhstan, utilizing both mobile and stationary air monitoring systems over a two-year period. The research focuses on tracking key air pollutants, namely carbon monoxide (CO), nitrogen dioxide (NO₂), particulate matter (PM2.5 and PM10), and sulfur dioxide (SO₂), providing a comparative assessment of seasonal trends and the sources of pollution, which include transportation, industrial emissions, and domestic heating during the cold season. The study emphasizes the significance of monitoring systems in urban environments to understand better the impact of air pollution on public health and the effectiveness of sustainable interventions. One of the major insights from this research is the comparison between seasonal variations in pollutant levels and the city's transition toward sustainable energy practices, such as increased gasification and the use of electric transportation, which has already demonstrated a positive impact on reducing emissions during peak heating periods. The results show that while Astana has improved air quality, air pollution remains a concern, especially in winter due to the increased use of solid fuel. This paper emphasizes the importance of real-time data from mobile sensors and suggests their wider use to complement stationary sensors for better monitoring. In addition to pollutant tracking, the study delves into the health implications of prolonged exposure to air pollutants, particularly in urban areas. The study concludes by advocating for expanded use of mobile monitoring systems and advanced data analytics to provide actionable insights for policymakers, urban planners, and public health officials
IMPLEMENTATION OF A BASE OF RULES FOR DIFFERENTIAL DIAGNOSIS OF CLINICAL AND HEMATOLOGICAL SYNDROMES BASED ON MORPHOLOGICAL CLASSIFICATION ALGORITHM
Get PDFThe evolving landscape of modern medicine underscores the growing importance of automating diagnostic processes. This advancement is not merely a convenience but a necessity to harness the full potential of technological progress, aiming to elevate research and clinical outcomes to new heights. Among the innovative strides in this field, the development of diagnostic systems based on morphological classification algorithms stands out. Such systems, rooted in comprehensive rule bases for differential diagnosis, promise to revolutionize the way we approach complex medical conditions. This paper introduces a cutting-edge system that epitomizes this evolution. Designed to harness the power of data analysis, it paves the way for groundbreaking research opportunities. At the heart of this system is a sophisticated set of rules derived from a morphological classification algorithm. This foundation enables the system to perform automated diagnoses of a wide array of clinical and hematological syndromes with unprecedented accuracy. A notable application of this technology is its ability to diagnose anemia by analyzing six distinct blood parameters and further categorize the anemia type based on biochemical criteria. The implications of such diagnostic capabilities are profound. By enabling the systematic collection and analysis of statistical data, the system facilitates in-depth research into the prevalence of diseases across different demographic groups. It aids in identifying disease patterns and supports preventive medicine efforts, potentially shifting the paradigm from treatment to prevention. This study not only highlights the system's capacity for enhancing diagnostic precision but also emphasizes its role as a catalyst for medical research and the improvement of healthcare delivery. The integration of such technologies into the medical field promises to enhance the quality of care, streamline diagnostic processes, and open new avenues for medical research, ultimately contributing to the advancement of global health standards
DEVELOPMENT OF AEROSPACE IMAGES PRELIMINARY PROCESSING METHOD FOR SUBSEQUENT RECOGNITION AND IDENTIFICATION OF VARIOUS OBJECTS
Get PDFNowadays, the application of hyperspectral images is vital for every section of the humanity life such as agrotechnical research for the field condition state and water security. This article presents a new lossless data compression algorithm focused on the processing of hyperspectral aerospace images. The algorithm takes into account inter-band correlation and difference transformations to effectively reduce the range of initial values. correlation allows you to find the best reference channel that defines the sequence of operations in the algorithm, which contributes to a significant increase in the compression ratio while maintaining high data quality. The practical implementation of the algorithm lies in the process of the transfer the lower size file with high efficiency for unmanned aerial vehicle and satellites to save more computational resources. This method demonstrates high computational efficiency and can be applied to various tasks that require efficient storage and transmission of hyperspectral images. The importance of processing hyperspectral data and the problems associated with their volume and complexity of analysis were described. Current approaches to data compression are considered and their limitations are identified, which justifies the need to develop new methods. The relevance and necessity of effective compression algorithms for aerospace applications is emphasized. An analysis of existing methods and algorithms for compressing hyperspectral data was carried out. Particular attention is paid to approaches that use cross-channel correlation and difference transformations. The effectiveness of current methods is evaluated and their shortcomings are identified, which serves as a justification for the development of a new algorithm. A developed lossless data compression algorithm based on the use of inter-band correlation and difference transformations was described. The stages of forming groups of channels and the selection of optimal compression parameters are considered in detail. The method of determining the reference channel, which sets the sequence of operations in the algorithm, which provides more efficient data compression, is especially noted. The advantages and possible limitations of the new approach, as well as its potential for practical use, are discussed. It was noted that the developed method successfully solves the problems associated with the volume of hyperspectral data, providing a high compression ratio without quality loss. The prospects for further development of the algorithm and its application in various fields of science and technology are discussed
DEVELOPING GAME THEORY-BASED METHODS FOR MODELING INFORMATION CONFRONTATION IN SOCIAL NETWORKS
Get PDFThis paper explores the essential dynamics of social networks, specifically examining the phenomenon of information confrontation among users. The goal of the research is the development of a novel simulation methodology that integrates game-theoretic principles with probabilistic techniques to provide a robust model for these interactions. The theoretical framework of the study is founded on the conceptualization of user conflicts as a strategic game between two players. The primary objective for each player in this game is to exert influence and control over as many nodes within the network as possible. To capture the essence of these strategic interactions, we have introduced an innovative algorithm that facilitates dynamic strategy adaptation. This algorithm is pivotal in allowing players to modify their decision-making processes in real-time, based on the continually changing conditions of the network. For practical implementation and validation of the methodology, we used the Facebook Researcher open dataset, with a particular focus on its Kazakhstani segment. This dataset provides a rich source of empirical data, reflecting diverse user interactions and network configurations, which are essential for testing the model. This approach stands out by offering significant improvements in computational efficiency and resource management. By dynamically tracking and updating the network's status, the proposed method reduces the computational resources required, thereby enhancing the scalability of the simulation. In comparing our methodology with other existing models in the field, it becomes evident that it not only matches but in several respects surpasses these methodologies in terms of flexibility. This study makes substantial contributions to the field of social network analysis by providing a sophisticated tool that can be effectively employed to navigate and analyze the complexities of information confrontation in digital social spaces
ANALYSIS OF THE IMPACT OF SHARDING ON THE SCALABILITY AND EFFICIENCY OF BLOCKCHAIN TECHNOLOGIES FOR THE CREATION OF INFORMATION-ANALYTICAL SYSTEMS FOR ENVIRONMENTAL MONITORING OF EMISSIONS INTO THE ENVIRONMENT
Get PDFThis study examines the impact of sharding on the scalability and efficiency of blockchain systems, specifically in the development of a complex of intelligent information and communication systems for environmental monitoring of emissions into the environment for decision-making in the context of carbon neutrality. Utilizing the Ikarus Network infrastructure, sharding was implemented on masternodes as a key technology to optimize transaction processing. Sharding enables the blockchain to be divided into multiple parallel chains, significantly increasing throughput and reducing the load on individual nodes. The results demonstrate a 70% increase in transaction processing speed, allowing the system to handle up to 5000 transactions per second, compared to the previous 3000 transactions per second. Network throughput increased by 50%, ensuring more efficient load distribution and stable operation even with high data volumes. Statistical analysis using ANOVA confirmed significant improvements in transaction processing speed, confirmation time, and resource usage post-sharding implementation. The F-value for transaction processing speed was 4567 with a P-value of 0.0001, indicating substantial improvements. Visual data analysis further confirmed these results, showing noticeable performance enhancements in the blockchain system. Distribution charts and histograms of transaction processing speed and confirmation time revealed an increase in the average number of transactions per second and greater system stability post-sharding. Sharding not only increased throughput but also enhanced system security by decentralizing data among shards, complicating potential cyberattacks. The study aimed to determine how sharding can improve the scalability and efficiency of blockchain systems. These improvements position the Ikarus Network as a promising solution for scalable and secure blockchain-based applications, especially for tasks related to carbon emission monitoring and management. These findings can underpin further study and the development of more efficient blockchain technologies