XVII International Scientific and Practical Conference «Science in modern society»

SCIENCE IN MODERN SOCIETY Proceedings of the XVII International Scientific and Practical Conference 7-8 October 2025 BEIJING. CHINA 2025 UDC 001.1 BBC 1 XVII International Scientific and Practical Conference « Science in modern society», October 7-8, 2025, Beijing. China. 78 p. ISBN 978-91-65424-41-8 DOI https://doi.org/10.5281/zenodo.17349992 Publisher: «SC. Scientific conferences» Main organization: Editor: Hans Muller Layout: Ellen Schwimmer The conference materials are in the public domain under the CC BYNC 4.0 International license. The publisher is not responsible for the materials published in the collection. All materials are provided in the author's edition and express the personal position of the participant of the conference. The sample of the citation for publication is Gugnin Aleksandr, Lisnievska Yuliia ANTI-ADVERTISING IN THE HOTEL BUSINESS // XVII International Scientific and Practical Conference «Modern science: fundamental and applied aspects», October 7-8, 2025, Beijing. China. Pp.911, URL: https://sconferences.com Contact information Website: https://sconferences.com E-mail: [email protected] Content Biological sciences Ulughbek Niyozov, Ilxom Mukumov, Dildora Xushvaqtova MEDICINAL PLANTS OF THE TURTKULYUKSAY STREAM, GELON VILLAGE, KASHKADARYA BOTANICAL AND GEOGRAPHICAL DISTRICT 3 Economic sciences Gafurova Dilshoda Ramazanovna ENHANCING THE INNOVATION CAPACITY OF HIGHER EDUCATION INSTITUTIONS 6 Batchayev M.A., Ryspekov B. K., Alymbekov S.K., N.A. Gumar CRYPTOCURRENCIES AS AN ALTERNATIVE FINANCIAL INSTRUMENT: A CRITICAL EVALUATION OF THEIR ADVANTAGES AND DISADVANTAGES 9 Konstantine Kacharava RISK MANAGEMENT PROSPECTS IN MEDICAL ORGANIZATIONS 16 Xiaoqi Cheng SOCIAL MEDIA CHALLENGES IN CONSUMER DECISION-MAKING 21 Geographical sciences Ulviyya Isgandarova THE PRINCIPAL CHARACTERISTICS OF THE SOILS OF THE NAKHCHIVAN AUTONOMOUS REPUBLIC AND THEIR CARTOGRAPHIC REPRESENTATION 24 Geological and mineralogical sciences Roza Mukiyatkyzy INTEGRATION OF AI AND GIS TECHNOLOGIES FOR DIGITAL PREDICTION AND MAPPING OF MINERAL DEPOSITS 31 Historical sciences Otar Nikoleishvili FROM THE HISTORY OF THE “JEWISH ANTI-FASCIST COMMITTEE” 39 Mathematical sciences Yakimova Nataliya MATHEMATICAL INTERPRETATION OF SIMPLE WORD-COMBINATIONS USING VECTOR LOGICAL ALGEBRA (ON THE EXAMPLE OF SLAVIC AND ROMANCE-GERMANIC GROUPS OF LANGUAGES) 46 Medical sciences Bruno Gorana PHYSIOTHERAPEUTIC TREATMENT OF LOW BACK PAIN, A REVIEW OF LITERATURE 49 Dorina Ruci, Vilson Ruci, Krisli Serani, Arsel Dizdari, Vasilika Gjika, Entela Shkodrani AUTOIMMUNE THROMBOCYTOPENIAS. DIFFERENTIAL DIAGNOSIS 54 Pedagogical sciences Fakhriya Aslanova THE ROLE AND IMPORTANCE OF PRESCHOOL EDUCATION IN AZERBAIJAN 56 Philological sciences Elmira Hasanova DIGITAL LEARNING PATH AS A PEDAGOGICAL TOOL IN ENGLISH LANGUAGE TEACHING: FROM CONCEPT TO PRACTICE 59 Technical sciences Anamika Raj, Noor Maizura Mohamad Noor, Rosmayati Mohemad, Noor Azliza Che Mat HYBRID HWOPSO-WOA FOR OPTIMAL FEATURE SELECTION IN DIABETIC RETINOPATHY PREDICTION 66 E.N. İbrahimova ALGORITHM SELECTION AND OPTIMIZATION FOR IMAGE-BASED CRACK DETECTION 67 Yesmagambetov Bulat-Batyr, Botayeva Saule, Kenzhebayeva Ulzhan, Adilet Makhanbetov, Sultanova Gulbanu, Tursumbayeva Adiya STATISTICAL INFORMATION PROCESSING IN RADIO ENGINEERING SYSTEMS 72 XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 3 Biological sciences Introduction. Medicinal plants are one of the most valuable natural resources, providing raw materials for pharmaceuticals, traditional medicine, and functional food production [8]. The Kashkadarya botanical-geographical district, particularly the Gissar mountain system and its valleys, is rich in endemic and rare medicinal flora [2,3]. Among them, the Turtkulyuksay stream, passing through Gelon village (Shahrisabz district), represents a unique microhsssabitat where traditional knowledge and natural diversity intersect. The aim of this study is to identify, classify, and evaluate medicinal plants growing in the Turtkulyuksay stream area, to analyze their ecological adaptation, and to determine their potential for pharmacological use [1]. The research was conducted during the vegetation periods of 2025. Field expeditions were carried out along the Turtkulyuksay stream and its surroundings at altitudes ranging from 2500 to 2550 m above sea level. Floristic lists were compiled based on route-method surveys. Species identification was performed using the "Flora of Uzbekistan" (1952–1987), complemented with recent taxonomic updates [4]. The study area is Turtkulyuksay, Gelon village, Shahrisabz district, Kashkadarya region of Uzbekistan (Figure 1). Figure 1. Satellite images of the study area. A. Location of Gelon village (Gilan) and its surrounding mountainous landscape. The red rectangle indicates the research site. B. Enlarged view of the Turtkuluksay stream basin with geographic coordinates (39°03′31″N – 67°31′01″E). The ecological groups of plants were determined according to their habitat preferences (hygrophytes, xerophytes, mesophytes) [5]. Ethnobotanical information was collected through structured interviews with local inhabitants of Gelon village. Results and discussion. In total, 27 species of higher plants belonging to 18 families were recorded in the Turtkulyuksay stream valley. The most represented families are: Asteraceae (Compositae) – 4 species (Artemisia dracunculus L., Tanacetum vulgare L., Ligularia thomsonii C.B.Clarke, Matricaria chamomilla L.), Apiaceae (Umbelliferae) – 3 species (Heracleum lehmannianum Bunge., Selinum carvifolia L., Prangos pabularia Lindl.), Lamiaceae (Labiatae) – 2 species (Mentha piperita L., Phlomis tuberosa L.), Cupressaceae – 2 species (Juniperus turkestanica Kom., Juniperus seravschanica Kom.), Polygonaceae – 2 species (Rumex confertus Willd., Polygonum aviculare L.), Hypericaceae – 2 species (Hypericum MEDICINAL PLANTS OF THE TURTKULYUKSAY STREAM, GELON VILLAGE, KASHKADARYA BOTANICAL AND GEOGRAPHICAL DISTRICT Ulughbek Niyozov Institute of Biochemistry, Samarkand State University named after Sharof Rashidov, Samarkand, Uzbekistan Ilxom Mukumov Institute of Biochemistry, Samarkand State University named after Sharof Rashidov, Samarkand, Uzbekistan Dildora Xushvaqtova Teacher at Mitti Academy Private School, Kitob District, Kashkadarya Region, Uzbekistan XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 4 perforatum L., Hypericum scabrum L.) and the remaining 12 families each have one plant. The distribution and proportion of each plant family is shown in Figure 2. Figure 2. Distribution of medicinal plant species by plant family in Gelon village, Turtkulyuksay The medicinal plants are distributed according to the ecological gradient of the valley: 1. Streamside hygrophytes (P. pabularia Lindl.) thrive in moist soils. 2. Slope xerophytes (J. turkestanica Kom., J. seravschanica Kom., R. confertus Willd., P. aviculare L., H. scabrum L.) 3. Meadow mesophytes (A. dracunculus L., T. vulgare L., L. thomsonii C.B.Clarke, M. chamomilla L., H. lehmannianum Bunge, S. carvifolia L., M. piperita L., Ph. tuberosa L., H. perforatum L. This ecological zonation reflects the microclimatic diversity of the Turtkulyuksay valley, which favors high medicinal plant diversity. Ethnobotanical significance. Local inhabitants of Gelon village possess rich traditional knowledge regarding medicinal plants: J. turkestanica Kom. – used in folk medicine for respiratory diseases, as disinfectant; wood for construction and incense, Berberis vulgaris L. – fruits rich in vitamin C; used for fevers, digestive disorders; bark for dyes, Ribes nigrum L. - berries used as food, source of vitamin C; leaves in folk medicine for colds and rheumatism, Quercus robur L. – bark used as astringent, for wound healing; wood for construction, M. piperita L. – widely used medicinal plant for digestion, colds; essential oil in perfumery, M. chamomilla L. – famous medicinal herb (chamomile tea); used for stomach, nervous system, skin. Conservation aspects. Some medicinal plants of the valley (Ephedra intermedia Schrenk & C.A.Mey., Ferula tenuisecta Korovin, and Crataegus turkestanica Pojark.) are listed in the Red Data Book of Uzbekistan due to habitat degradation and overharvesting. Therefore, conservation strategies such as controlled harvesting, cultivation, and community-based management are essential [6]. Conclusion. The Turtkulyuksay stream valley in Gelon village represents a rich reservoir of medicinal plants, comprising more than 40% of the total recorded flora. The dominance of families such as Asteraceae, Apeaceae and Lamiaceae highlights the ecological and pharmacological value of the area. Traditional knowledge preserved by local inhabitants contributes significantly to the sustainable use of these resources. Future perspectives include: establishing ex situ cultivation plots for rare medicinal plants, promoting eco-tourism and ethnobotanical education, conducting phytochemical screening of endemic species for novel drug discovery [7]. Thus, the Turtkulyuksay stream valley is not only a natural heritage site but also a promising resource for the pharmaceutical industry and ethnobotany of Uzbekistan. References 1. Flora of Uzbekistan, Vols. I–IX. Tashkent: Fan, 1952–1987. 2. Karimov, F.Kh., & Tojiboev, K.Sh. (2010). Medicinal plants of Uzbekistan. Tashkent: Universitet. 3. Red Data Book of the Republic of Uzbekistan. (2019). Tashkent. 4. Khassanov, F.O. (2021). Biodiversity of the Gissar mountain system. Journal of Central Asian Botany, 7(2), 56–69. 5. WHO. (2020). Traditional, complementary and integrative medicine. Geneva 6. Mamadalieva N.Z. et al. Aromatic Medicinal Plants of the Lamiaceae Family from Uzbekistan: Ethnopharmacology, Essential Oils Composition, and Biological Activities. Medicines (MDPI). DOI: 10.3390/medicines4010008. 7. D. Egamberdieva, D. Jabborova. “Medicinal Plants of Uzbekistan and Their Traditional Uses”. Vegetation of Central Asia and Environs (Springer). DOI: 10.1007/978-3-319-99728-5_8. 432222 12 0 2 4 6 8 10 12 14 Number of Plant Species Families XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 5 8. Eshonkulov A. H. et al. Ethnobotanics of Certain Medicinal Plants of Bukhara Region (Uzbekistan). American Journal of Plant Sciences, DOI: 10.4236/ajps.2022.133024. XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 6 Economic sciences In the era of the Fourth Industrial Revolution, the role of higher education has undergone a profound transformation from being a transmitter of established knowledge to becoming a generator of new knowledge and innovation. Modern universities are increasingly recognized as innovation hubs, where education, science, and entrepreneurship converge to create new ideas, technologies, and solutions that shape national and global development. The effectiveness of higher education systems is now measured not only by teaching quality or graduate employment rates but also by their ability to produce research-driven innovation and to contribute to the knowledge economy. The development of innovative activity in higher education represents a multidimensional and systemic process that encompasses scientific research, creative thinking, digital transformation, and entrepreneurial behavior among faculty and students. It reflects a paradigm shift in which universities move from traditional, lecture-centered institutions to innovation-oriented ecosystems that interact dynamically with industry, government, and society. In this context, universities serve as both incubators of innovative ideas and accelerators of technological and socio-economic progress. From a theoretical standpoint, innovation in higher education can be interpreted through the lens of Schumpeter’s (1934) innovation theory, which identifies innovation as the primary driver of economic and institutional renewal. In the educational context, this renewal manifests through the introduction of new teaching methodologies, digital tools, research-based curricula, and interdisciplinary collaboration. Moreover, the Triple Helix Model developed by Etzkowitz and Leydesdorff (2000) emphasizes that universities, industries, and governments must act in synergy to build a sustainable national innovation system. This model is particularly relevant for emerging economies such as Uzbekistan, where higher education reform is a key component of national modernization strategies. Despite significant progress in the global education sector, many universities especially in developing countries continue to face structural, organizational, and financial barriers that hinder innovation. Limited research funding, outdated curricula, insufficient digital infrastructure, and weak linkages between academia and industry restrict the ability of universities to operate as engines of innovation. In addition, innovation in education is not limited to technology adoption but also involves cultural and institutional transformation creating an environment that encourages experimentation, risk-taking, and interdisciplinary collaboration. The development of innovative activity within universities is also closely associated with human capital formation. Academic staff must be equipped with research competence, pedagogical creativity, and entrepreneurial skills, while students should be trained to think critically, act independently, and engage in innovation-oriented projects. The university’s innovation potential therefore depends on both its structural capacity (infrastructure, funding, and governance) and its cultural capacity (values, motivation, and openness to change). In Uzbekistan, the past decade has seen major reforms in the field of higher education aimed at integrating the national system into the global educational space. The adoption of the “Strategy for the Development of Higher Education 2030”, the creation of research and innovation clusters, and the expansion of academic autonomy have set the foundation for innovation-oriented development. However, ENHANCING THE INNOVATION CAPACITY OF HIGHER EDUCATION INSTITUTIONS Gafurova Dilshoda Ramazanovna PhD., Acting Professor, Department of Management and Marketing, Kimyo International University in Tashkent, Republic of Uzbekistan Abstract This paper examines the role of higher education institutions as key drivers of innovation in the modern knowledge economy. It analyzes theoretical foundations, organizational mechanisms, and strategic factors influencing the development of innovative activity in universities. The study highlights the importance of entrepreneurial governance, digital transformation, and university industry collaboration for building sustainable innovative ecosystems. Keywords: Higher education, innovative activity, university management, entrepreneurship, digital transformation, research ecosystem. XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 7 realizing this potential requires not only regulatory changes but also the transformation of the university’s internal processes research management, technology transfer, digitalization, and partnership with industry. Global trends demonstrate that universities which successfully institutionalize innovative activity share several features: they operate with high academic freedom, possess flexible management systems, and maintain diversified sources of funding. Moreover, they promote innovation culture through entrepreneurship education, digital learning platforms, and collaborative research ecosystems. For countries in transition, developing such models is both a challenge and a strategic necessity. The contemporary understanding of innovation in higher education (HE) draws from classic economic and organizational theories. Schumpeter’s (1934) theory of creative destruction positions innovation as a systemic process of renewal that reconfigures resources and routines an idea that translates to universities as they re-design curricula, research agendas, and governance to generate new value. Drucker (1985) reframes innovation as a disciplined managerial practice, aligning well with institutional mechanisms research strategy, performance incentives, and portfolio management used by HEIs to scale innovative activity. On the diffusion side, Rogers (2003) stresses the roles of communication networks, perceived attributes of innovations, and change agents, all salient in academic departments and interfaculty collaborations. In knowledge production, the shift from Mode 1 (discipline-bound) to Mode 2 (transdisciplinary, application-oriented) research (Gibbons et al., 1994) helps explain the growing salience of mission-driven, problem-focused projects, living labs, and challenge-based learning within universities. The knowledgecreating firm perspective (Nonaka & Takeuchi, 1995) with the SECI model (socialization, externalization, combination, internalization) has been frequently repurposed for universities, highlighting how tacit and explicit knowledge circulate through labs, centers, and technology transfer offices. Entrepreneurial university literature emphasizes organizational adaptations that enable innovation. Clark (1998) identifies five transformation pathways—strengthened steering core, expanded development periphery, diversified funding base, stimulated academic heartland, and an integrated entrepreneurial culture. Complementary analyses examine how rankings/regulatory pressures reshape institutional behavior (Hazelkorn, 2015) and how managerial reforms alter academic work (Bleiklie, 2018). The debate on academic capitalism (Slaughter & Rhoades, 2004) underscores opportunities and risks of market-oriented strategies: while third-stream activities and IP income may stimulate innovation, they can also skew incentives away from public-good science. The Triple Helix model (Etzkowitz & Leydesdorff, 2000; Etzkowitz, 2008) conceptualizes innovation as an outcome of recursive interactions among universities, industry, and government, positioning HEIs as both knowledge producers and entrepreneurial actors. Extensions to a Quadruple Helix integrate civil society and media-based culture as co-producers of innovation (Carayannis & Campbell, 2009). Empirical studies document pathways contract research, licensing, and spin-offs through which universities influence regional innovation systems (Perkmann et al., 2013; Wright, Siegel, & Mustar, 2017), and the role of science parks and incubators in amplifying spillovers (Link & Scott, 2003). The open innovation lens (Chesbrough, 2003) further explains why HEIs increasingly orchestrates porous boundary arrangements consortia, standard-setting bodies, and shared IP platforms. A substantial stream analyzes the organization and performance of technology transfer from TTO structures and incentive schemes to IP regulation and faculty engagement (Siegel & Wright, 2015). At the human-capital level, entrepreneurship education is linked to opportunity recognition, entrepreneurial intention, and venture creation (Fayolle & Gailly, 2008; Rae, 2010). Capability building is cumulative and path-dependent: world-class institutions combine doctoral training, post-doc pipelines, interdisciplinary centers, and PhD-to-startup pathways (Altbach & Salmi, 2011; Marginson, 2016). International evidence indicates that diversified revenue, academic freedom, and robust research infrastructure are persistent correlates of innovative university performance (OECD, 2023; World Bank, 2023; UNESCO, 2022). Digital transformation functions as both a catalyst and infrastructure for innovative activity. Scholarship highlights the strategic use of learning analytics, digital repositories, and virtual/remote labs to expand research-teaching synergies (Bates, 2015; Selwyn, 2016; Weller, 2020). Systematic reviews show accelerating adoption of AI in higher education for adaptive learning, early-warning systems, and assessment (Zawacki-Richter et al., 2019), while cautioning about ethics, bias, and governance. These developments align with Pasteur’s Quadrant (Stokes, 1997), where use-inspired basic research links disciplinary inquiry with societal application now increasingly mediated by data infrastructures. Developing innovative activity within higher education institutions (HEIs) demands a comprehensive transformation of governance, research culture, and digital infrastructure. To build sustainable XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 8 innovation ecosystems, universities should establish innovation governance frameworks that link academic units with industry, government, and society. Dedicated innovation councils or technology-transfer offices can coordinate project funding, intellectual-property management, and startup incubation. A shift toward entrepreneurial education embedding creativity, design thinking, and research-to-market skills into all levels of curricula is essential to align graduate competencies with innovation-driven labor markets. Equally vital is the expansion of digital and research infrastructure: smart campuses, open-access repositories, and data-driven management systems that support experimentation and cross-disciplinary collaboration. The state and private sector must jointly invest in competitive grant schemes, publicprivate partnerships, and digital learning platforms to ensure continuity of research and innovation processes. Moreover, universities should internationalize their innovative ecosystems through joint doctoral programs, global research networks, and participation in transnational innovation clusters to internalize best practices and attract high-level talent. In conclusion, the advancement of innovative activity in higher education is both a strategic driver of national competitiveness and a core element of sustainable socio-economic development. Universities can no longer function merely as teaching organizations; they must evolve into integrated knowledge enterprises that generate, transfer, and commercialize intellectual output. The synergy of academic autonomy, entrepreneurial governance, and digital transformation forms the cornerstone of this evolution. Strengthening human capital through continuous professional development, fostering an innovation-oriented academic culture, and embedding ethical and inclusive principles in innovation policies will ensure that higher education becomes a catalyst for technological progress and social well-being. Thus, innovation in higher education should be viewed not only as an institutional reform agenda but as a national innovation mission one that transforms knowledge into value, research into development, and education into long-term prosperity. References 1. Altbach, P. G., & Salmi, J. (Eds.). (2011). The road to academic excellence: The making of worldclass research universities. Washington, DC: World Bank. 2. Bates, A. W. (2015). Teaching in a digital age: Guidelines for designing teaching and learning. Vancouver, BC: BCcampus. 3. Bleiklie, I. (2018). New public management or neoliberalism, higher education. In P. Teixeira & J. C. Shin (Eds.), Encyclopedia of international higher education systems and institutions (pp. 1–7). Dordrecht: Springer. 4. Carayannis, E. G., & Campbell, D. F. J. (2009). 'Mode 3' and 'Quadruple Helix': Toward a 21st century fractal innovation ecosystem. International Journal of Technology Management, 46(3/4), 201– 234. 5. Chesbrough, H. (2003). Open innovation: The new imperative for creating and profiting from technology. Boston: Harvard Business School Press. 6. Clark, B. R. (1998). Creating entrepreneurial universities: Organizational pathways of transformation. Oxford: Pergamon. 7. Drucker, P. F. (1985). Innovation and entrepreneurship: Practice and principles. New York: Harper & Row. 8. Etzkowitz, H. (2008). The Triple Helix: University–industry–government innovation in action. London: Routledge. 9. Etzkowitz, H., & Leydesdorff, L. (2000). The dynamics of innovation: From National Systems and “Mode 2” to a Triple Helix of university–industry–government relations. Research Policy, 29(2), 109–123. 10. Fayolle, A., & Gailly, B. (2008). From craft to science: Teaching models and learning processes in entrepreneurship education. Journal of European Industrial Training, 32(7), 569–593. 11. Ramazanovna, G. D. (2024). ANALYSIS OF THE REFORMS IMPLEMENTED IN THE HIGHER EDUCATION SYSTEM OF THE REPUBLIC. Scientific Journal of Actuarial Finance and Accounting, 4(03), 249255. 12. Ramazanovna, G. D. (2023). SOME ASPECTS OF INNOVATIVE MANAGEMENT IN HIGHER EDUCATION. BBC, 19. 13. Ramazanovna, G. D. (2020). Methodology for Evaluating the Management of Innovative Processes in the Field of Information and Communication Technologies. Webology, 17(1), 365-376. XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 15 9. Sky.pro. Total number of cryptocurrencies in the world: complete statistics and market dynamics [Electronic resource]. – Available at: https://sky.pro/wiki/money/skolko-vsego-kriptovalyut-v-mirepolnaya-statistika-i-dinamika-rynka (accessed: 12.09.2025). 10. Ushakov D.S., Podolskaya T.V., Tomashevskaya L.A., Sysoeva A.A. International cryptocurrency market and distinguishing features of its development [Electronic resource] // Public and Municipal Administration. Scientific Notes, 2019. – Pp. 85–89. – DOI: 10.22394/2079-1690-2019-1-4-85-89. – Available at: https://cyberleninka.ru/article/n/international-cryptocurrency-market-and-distinguishing-features-of-its-development (accessed: 28.09.2025). XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 16 Risk management in healthcare organizations is a comprehensive process that identifies, analyzes, and mitigates threats that could impact patient safety, staff performance, and the sustainability of the entire healthcare system. With the rapid advancement of medical technologies, the increasing complexity of clinical processes, and increasing demands on service quality, risk management is becoming an integral part of the strategic management of healthcare institutions [1]. Modern approaches to risk management are based on international standards such as ISO 31000, as well as national regulations adapted to the specific needs of individual countries. Implementing these standards allows for the systematization of efforts to prevent adverse events, minimize the impact of human error, and ensure the reliability of medical processes. However, experience shows that implementing risk management methodologies in healthcare institutions is fraught with several challenges related to organizational characteristics, human resources, and the level of digital transformation in healthcare [2]. A crucial aspect of risk management in medicine is identifying potential threats. Errors by medical personnel, equipment malfunctions, inadequate specialist training, medication shortages, and disruptions in interdepartmental communication can all lead to negative consequences for patients and a decline in the quality of medical care. Therefore, one of the top priorities for managing medical organizations is to develop an effective risk identification system, including regular monitoring and situational analysis [3]. Experience shows that most risks in medicine are systemic in nature. This means that adverse events rarely occur in isolation, but are the result of a combination of factors. For example, medication errors can be caused not only by physician inattention, but also by staff shortages, employee overload, insufficient integration of information systems, or inadequate regulations [4]. Therefore, for effective RISK MANAGEMENT PROSPECTS IN MEDICAL ORGANIZATIONS Konstantine Kacharava The Doctoral Student of Caucasus International University Tbilisi, Georgia Abstract This article examines risk management in healthcare facilities, examining modern methods and international standards, including ISO 31000. Key approaches to identifying, assessing, and mitigating risks affecting patient safety and the performance of healthcare personnel are analyzed. The challenges faced by healthcare organizations are highlighted, and practical solutions for implementing risk management systems are proposed. Special attention is given to Georgia's experience, national standards, and prospects for their further development. The importance of an integrated approach to risk management, including the use of expert assessments and specialized methodologies such as HAZOP and FMEA, is emphasized to ensure a high level of safety and quality of medical care. The article highlights that most risks in medicine are systemic in nature. It concludes that adverse events rarely happen in isolation but are usually the result of multiple factors. An example provided is medication errors, which can be caused not only by physician inattention but also by staff shortages, employee overload, inadequate integration of information systems, or insufficient regulation. The work demonstrates that effective risk management requires not only recording individual errors but also analyzing their causes and developing systemic prevention strategies to mitigate them. It emphasizes that risk assessment is a crucial part of the management process, as different threats have varying probabilities of occurrence and various levels of impact on healthcare operations. The article demonstrates that both qualitative and quantitative risk assessment methods are employed globally. Some organizations depend on expert assessments based on the experience of physicians and administrators, while others use mathematical models to estimate the likelihood of specific events. The article concludes by noting that modern medical centers actively implement automated monitoring systems that analyze patient data, staff workload, equipment condition, and other parameters, allowing them to predict potential failures and take preventative measures. Once risks are identified and assessed, they are mitigated. Keywords: Risk management, medicine, safety, standards, threat analysis. XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 17 risk management, it is essential not only to record individual errors but also to analyze their root causes and develop prevention mechanisms at the systemic level. Risk assessment is another critical step in the management process. Different threats have different probabilities of occurrence and varying degrees of impact on the operations of a medical institution. Both qualitative and quantitative risk assessment methods are used worldwide. Some organizations rely on expert assessments based on the experience of physicians and administrators, while others employ mathematical models to predict the likelihood of certain events. Modern medical centers are actively implementing automated monitoring systems that analyze patient data, staff workload, equipment status, and other parameters, enabling them to predict potential disruptions and implement preventative measures. After identifying and assessing risks, mitigation follows. Different strategies are employed depending on the nature of the threats. Some risks can be mitigated by modifying organizational processes, such as implementing electronic medication management systems, automating document management, or standardizing medical protocols. In other cases, staff engagement is required, including professional development, the development of precise instructions, training, and simulation exercises. Technical support plays a special role—the use of modern technologies, robotic systems, and artificial intelligence algorithms can reduce the likelihood of medical errors and improve diagnostic accuracy. Georgian healthcare institutions are also adapting international risk management practices to local realities. The country has already implemented national standards, such as ISO 31000:2020, which regulates key approaches to risk management. However, practice shows that successful implementation of these standards requires not only regulatory changes but also a thorough restructuring of management processes within the organizations themselves. Specifically, it is essential to develop a safety culture among medical personnel, encourage the active participation of doctors and nurses in identifying and eliminating threats, and establish effective internal control mechanisms to ensure patient safety. A feedback system plays a crucial role in risk management. Many countries utilize specialized platforms where healthcare workers can report potential threats and errors anonymously, without fear of administrative sanctions. This approach not only allows for the recording of emerging issues but also enables the analysis of trends, identification of the most vulnerable processes, and the development of effective preventive measures. Georgia does not yet have a comprehensive system for collecting and analyzing information on medical errors, but the first steps in this direction are already being taken. Another essential element is the integration of risk management with the healthcare quality assurance system. In developed countries, these processes are closely linked: mortality data, readmissions, patient satisfaction, and other indicators are analyzed to identify systemic problems. In Georgia, this area is only beginning to develop; however, the digitalization of healthcare opens up opportunities for implementing analytical tools that enable real-time assessment of healthcare organization performance. Risk management in medicine is impossible without adequate funding. In developed countries, significant budgets are allocated for this purpose, enabling the development of automated monitoring systems, extensive staff training, and the implementation of innovative technologies. In Georgia, funding for healthcare organizations remains limited, creating additional barriers to the effective implementation of risk management. Therefore, the search for alternative funding sources, including grants, international programs, and private investment, is critical. Despite all the difficulties, the development of a risk management system in healthcare is a necessity dictated by modern realities. The growing number of patients, the increasing complexity of medical procedures, and high public expectations regarding the quality of medical care necessitate that healthcare institutions implement modern risk analysis and management methods. Experience from other countries shows that a systematic approach to risk management not only reduces the incidence of adverse events but also improves the overall level of medical safety, making the care process more predictable and manageable. Modern risk management methods in medical institutions are a crucial area of healthcare development, as they ensure patient safety, minimize professional risks for staff, and enhance the overall efficiency of medical organizations. Risk management is a set of measures designed to identify, analyze, and mitigate potential threats that may arise during the provision of medical services. In the complex and rapidly changing realities of healthcare, risk management is becoming an integral part of the daily practice of medical institutions [1]. Risk management principles and standards are grounded in international experience, drawing on recommendations from organizations such as the World Health Organization (WHO), the International XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 18 Organization for Standardization (ISO), and other professional societies. One of the fundamental documents is the ISO 31000 standard, which establishes the key stages of risk management, including risk identification, assessment, management, and monitoring. This standard is universal and adaptable to various conditions, including those found in medical institutions. Risk management in medicine covers a wide range of areas, from preventing human errors to implementing quality control systems for equipment and processes. This approach is based on the principle of the systems approach, which requires analyzing and considering multiple interrelated factors. For example, human errors, which often lead to undesirable consequences, are usually caused by systemic failures, such as a lack of resources, ineffective communications, or inadequate employee training. Consequently, risk management requires a comprehensive approach that includes both improving internal processes and developing external collaborations [2]. To achieve the study's objectives, a comprehensive search and analysis of the scientific literature on risk management in healthcare organizations was conducted. This review demonstrated significant interest in the topic under study from the global scientific community. In the initial phase of the study, publications presented in domestic and international peer-reviewed journals were examined. The primary focus was on analyzing risk management approaches in healthcare organizations, assessing the risks associated with medical personnel's professional activities, and evaluating tools for measuring patient safety culture used globally. Official information sources were also consulted, including departmental websites from Georgia and other countries, covering risk management and the development of patient safety culture in healthcare organizations. Risk management in healthcare facilities encompasses a wide range of activities, including strategies, processes, methods, and actions designed to prevent adverse events and incidents. These approaches encompass clinical and non-clinical risks that may arise during the diagnosis, treatment, rehabilitation, and care of patients. They help staff at various levels identify, analyze, assess, and minimize risks affecting both patient and employee safety [3]. In this study, the term “risk” refers to any undesirable event that combines potential harm and the likelihood of its occurrence. This definition aligns with contemporary risk management concepts and standards. International and national standards, including Georgian standards, offer generally accepted risk management principles and methods, providing a basis for their practical application in various fields. These standards aim to fully address potential threats that may hinder the achievement of established goals. Complete risk identification is of strategic importance, as an undetected risk can significantly distort the analysis of the current situation. Risk management standards recommend using questionnaires as one of the most effective methods for identifying threats in the early stages of the process. This sociological method allows for the identification of general trends and the creation of a unified database for analyzing the risk structure across departments [4]. Since 2011, Georgia has had several national risk management regulations in place, including ISO 31000:2020, “Risk Management - Guidelines,” IEC 31010:2020, “Risk Management - Risk Assessment Methods,” and ISO 73:2010, “Risk Management - Vocabulary.” These standards provide flexibility of application and can be adapted to various industries. For example, ISO 31000 offers recommendations suitable for risk management in any organization operating within a quality management system [5]. The universality of risk management standards allows for the specific characteristics of each organization to be taken into account. For example, certain aspects of medical personnel's work add unique characteristics to risk assessment. At the same time, the use of unified approaches promotes the unification of processes and the development of a common risk management culture. The study of risk management in healthcare organizations is based on a combination of international experience, national standards, and practical methods. This enables the effective identification, analysis, and mitigation of potential threats, improving patient safety and staff performance. Risk management in healthcare requires considering a multitude of potential threats that can impact the quality of medical care and patient safety. Medical activities are associated with numerous complexities that necessitate a systematic approach to identifying, analyzing, and mitigating risks. According to international and national standards, risk management involves several stages. The first step is risk identification, which consists of identifying all possible sources of threats, classifying them, and describing their characteristics. Georgian National Standard ISO 73-2010 defines risk identification as a process that enables the compilation of a threat inventory and ensures its complete understanding. This is an important step, as missed risks can lead to significant problems in the future [6]. After risk identification, they are assessed. This stage involves analyzing the likelihood of events, their causes, and possible consequences. This approach enables the early identification of threats that XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 19 require immediate action and those that can be addressed later. The assessment process integrates several components: identification, analysis, and subjective perception of threats. Risks can be analyzed at the organizational, departmental, or process-specific levels. Various methodologies are used for risk assessment, including both quantitative and qualitative approaches. For example, in some cases, the emphasis is on the likelihood of an event, while in others, the scale of potential damage is emphasized. The use of various methodologies allows for the assessment process to be tailored to the specific objectives and operating conditions of a healthcare organization. Expert assessment is one of the most effective approaches to risk analysis. This method relies on the opinions and experience of specialists who make judgments about potential threats and ways to minimize them. Involving competent experts allows for consideration of various aspects of the problem, which is especially important in strategically significant issues. Collaboration among specialists facilitates more accurate risk assessment and improved management. Modern approaches to risk analysis also include the use of specialized methodologies, such as HAZOP. 1 and FMEA 2 . Each of these methodologies has its own characteristics and limitations. For example, HAZOP requires extensive information about the subject of study, while FMEA is more suitable for analyzing threats and their consequences, but is not always applicable for retrospective assessment. The choice of an appropriate method depends on the specific conditions and objectives of the study [7]. Therefore, risk management in healthcare organizations requires a comprehensive approach that includes systematic threat identification, analysis, and the development of mitigation measures. Effective implementation of these steps helps create a safe environment for both patients and medical personnel, while also improving the quality of services provided. References 1. Accreditation Handbook/J. Brennan, I. Gocdegebuure, J. Sban, D. Westerbeijden et al. // Higher Education in Europe. – UNESCO European Centre for Higher Education, 1993. – Vol. XVIII, N 2. – P. 332. 2. Adams, H. W. Integriertes Management System für Sicherheit und Umweltschutz / H. W. Adams // Hanser Verlag. – Munchen – Wien, 1995. – P. 342. 3. “Baldrige Index” Consistently Outperforms the S&P 500 // Fact Sheet. – Gaithersburg, Maryland: NIST, 2001. – 10 April. – 49 p. 4. Bank, J. Quality and total quality management [ელექტრონული რესურსი] / J. Bank. – Bloomsbury Business Library – Information Sources, 2007. – P. 98. http://web.ebscohost.com/bsi/detail?vid=7&hid=120&sid=c9fee0c2-16ce4d55b96b686663570cc9%40SRCSM2&bdata=JnNpdGU9YnNpLWxpdmU%3d#db=bth&AN=26682650. 5. BSI PAS 56 Guide to Business Continuity Management [ელექტრონული რესურსი]: http://agsc.org.uk/uploads/docs/PAS56guidetoBCM.pdf. 6. Bush D., The deming prize and the baldrige award: how they compare/ D. Bush // Quality Progress. – 1989. – Vol. 22, № 1. – P. 254. 7. Cassel, C. Customer satisfaction and retention in transitional economies / C. Cassel, L. Murguletz // SSE S.-Pb Research Paper. – St. Petersburg, 2009. – 56 p. 8. Coleman, S. Where next for ISO 9000 companies?S. Coleman, Douglas // The TQM Magazine. – 2003. – Vol. 15, N 2. – P. 110. 9. Criteria for performance excellence//Baldrige National Quality Program. – 2002. – 58 p. 10. Dale, B. G. Managing quality; 3rd ed. / B. G. Dale. – Oxford: Blackwell Publishers, 1999–376 p. 11. Deming W. E. Out of the crisis / W. E. Deming. – Cambridge, MA: Massachusetts Institute of Technology, Center for Advanced Engineering Study, 1988. - 236 p. 12. Donabedian A. Criteria, norms and standards of quality: what do they mean? / A. Donabedian // Amer. J. Publ. Hlth. – 1981. – V. 71. 13. Donabedian A. The quality of medical care methods for issuing and monitoring the quality of care research and for quality assurance programs / A. Donabedian // Science. – 1978. – N 20. – P. 856– 1 HAZOP (Hazard and Operability Study) is a structured team analysis method designed to identify potential hazards and operational problems in process systems and equipment during the design or operation of industrial facilities, particularly in the chemical, oil and gas, and other highrisk industries. The purpose of this method is to identify potential deviations from design parameters, their causes, and consequences, thereby preventing accidents and enhancing the safety and reliability of production operations. 2 FMEA (Failure Mode and Effects Analysis) is a structured method for identifying, analyzing, and preventing potential defects in products and processes. Its purpose is to explore potential failure modes, their causes, and consequences, to take proactive action and reduce risks. FMEA is used to improve quality, reliability, and safety, as well as to reduce costs at all stages of the product or process lifecycle. XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 20 864. 14. Douglas A. The case for ISO 9000 / A. Douglas, S. Coleman // The TQM Magazine. 2003.Vol. 15. N5.-P. 316-324. 15. Feigenbaum, A. V. Total quality control; 3rd ed. / A. V. Feigenbaum. – N.- Y.: McGraw–Hill Book Co, 1983. – 253 p. 16. Group, ACE. n.d. 11 Critical Risks Facing the Healthcare Industry - Sponsored Content by ACE Group: Risk & Insurance. Accessed June 21, 2022. https://riskandinsurance.com/11-criticalrisks-facingthe-healthcare-industry/. 17. Ishikawa, K. What is total quality control? / K. Ishikawa // The Japanese Way. – N.-Y. : Prentice– Hall Inc., 1985. – 279 p. 18. It’s time to review your patient's safety culture https://hci.care/review-your-patient-safety-culture/ 19. Juran, J. M. Juran's quality control handbook / J. M. Juran, F. Gryna. – N.-Y.: McGraw–Hill Book Co, 1988. – 195 p. 20. Martinez-Costa, M. Effects of ISO certification on firms' performance: A vision from the market / M. Martinez-Costa, A.R. Martinez-Lorente // TQM & Business Excellence. – 2003. – Vol. 14, N 10. – P. 1179–1191. 21. Murrау, С. J. L. A framework for assessing the performance of health systems. Theme Papers / С. J. L. Muггау, J. Frеnk // Bulletin of the World Health Organization. – WHO, 2000. – Vol. 78. – P. 717– 724. 22. Price J., and Forrest J. Safety management systems. Practical Airport operations, safety, and emergency management. https://doi.org/10.1016/B978-0-12-800515-6.00005-6. 10.02.2025. 23. Slawomirski L, Auraaen A, Klazinga N. Organisation for Economic Co-operation and Development: Paris, France; 2017. https://psnet.ahrq.gov/issue/economics-patient-safetystrengthening-valuebased-approach-reducing-patient-harm-national 20.02.2025 24. The economics of patient safety in primary and ambulatory care: Flying blindhttps://read. ilibrary.org/social-issues-migration-health/the-economics-of-patient-safety-in-primary-andambulatorycare_baf425ad-en 03.02.2025 25. Watson P., Maslow D., Chileshe N., N.Y , 2004 26. Watson, P. A new model for obtaining sustainable competitive advantage/ P. Watson, N. Chileshe, D. Maslow // Construction Industry Development 2nd Postgraduate Conference (Cape Town, South Africa, 10–12 Oct. 2004). – 2004. – P. 129–141 27. Webair H., Al Assani S., Al Haddad R., Al Shaeeb W., Bin Selm M., and Alyamani A. Assessment of patient safety culture in primary care setting, Al-Mukala, Yemen // BMC Family Practice. – 2015. Vol. 16 №1. – P. 136. https://doi.org/10.1186/s12875-015-0355-1. 28. Why do errors happen? - To err is human. NCBI Bookshelf https://www.ncbi.nlm.nih.gov/books/NBK225171/#. 10.01.2025. XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 21 1. Introduction However, this transformation introduces complexities to the consumer decision-making process, traditionally structured into five stages—problem recognition, information search, evaluation of alternatives, purchase decision, and post-purchase evaluation (Kotler & Keller, 2022). Social media blurs the boundaries of these stages through continuous exposure to advertising, influencer endorsements, and algorithmic recommendations. The line between information and persuasion becomes indistinct. This study seeks to explore the inherent problems in this new paradigm, focusing on how the mechanisms of social media—virality, algorithmic personalization, and social validation—create new barriers to rational decision-making. Specifically, it addresses three central research questions: (1) How does the abundance of information on social media affect consumers’ ability to assess authenticity? (2) In what ways do online communities and algorithms reinforce cognitive biases and group polarization? (3) How do privacy violations and data exploitation undermine consumer trust? By addressing these questions, this paper aims to contribute to a more nuanced understanding of consumer vulnerability in digital environments and propose mechanisms for sustainable, ethical digital marketing practices. 2. Problems of Information Authenticity One of the most pervasive challenges in the social media ecosystem is the decline in information authenticity. Unlike traditional advertising channels, social media allows virtually anyone to create and disseminate content. This democratization of communication, while empowering, often leads to the proliferation of misinformation and deceptive marketing practices. Empirical research by Kim and Johnson (2023) shows that over 58% of users encounter misleading product information on social platforms monthly. The phenomenon of “influencer marketing fraud” further exacerbates this issue, where individuals with fake followers or undisclosed sponsorships distort consumer perceptions. The viral spread of low-quality or biased information creates what Wardle and Derakhshan (2017) term “information disorder,” where falsehoods circulate faster and wider than verified content. Algorithmic curation amplifies these effects. Social media platforms, optimized for engagement, prioritize emotionally charged and attention-grabbing content. As a result, consumers are exposed to information that appeals to affective responses rather than cognitive reasoning. A study by Vosoughi et al. (2018) found that false news spreads six times faster than factual news due to its novelty and SOCIAL MEDIA CHALLENGES IN CONSUMER DECISION-MAKING Xiaoqi Cheng Al-Farabi Kazakh National University, Almaty, Kazakhstan Abstract In the digital age, social media has transformed the traditional paradigms of consumer decision-making. Once characterized by a linear sequence of rational evaluations, the consumer journey is now dynamic, interactive, and heavily mediated by algorithms and online communities. This paper critically analyzes the challenges arising from the influence of social media on consumer decision-making, focusing on three primary problem domains: information authenticity, group polarization and cognitive bias, and privacy and data security. Drawing on recent empirical studies between 2021 and 2025, this research identifies how algorithm-driven personalization, influencer marketing, and user-generated content simultaneously empower and manipulate consumers. The findings reveal that while social media enhances market accessibility and participatory communication, it also undermines consumer autonomy through misinformation, emotional contagion, and invasive data profiling. The study emphasizes the need for balanced governance mechanisms that integrate algorithmic transparency, fact-checking systems, and privacy regulation. The research contributes to the growing field of behavioral economics and digital consumer studies by outlining a conceptual framework that connects digital media affordances with shifts in consumer cognition and trust. Keywords: Social media; consumer behavior; decision-making; information authenticity; cognitive bias; data privacy; digital marketing; behavioral economics XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 22 emotional salience. This pattern undermines consumers’ capacity to make informed decisions and erodes trust in digital marketplaces. Additionally, the blending of organic and sponsored content reduces transparency. Platforms like TikTok and Instagram increasingly feature native advertising formats that mimic user-generated posts, making it difficult for consumers to distinguish authentic recommendations from paid promotions. This “authenticity illusion” creates a cognitive burden, forcing consumers to continuously evaluate the credibility of sources and messages. 3. Group Polarization and Cognitive Bias The second major challenge arises from social media’s tendency to amplify cognitive biases and foster group polarization. Unlike traditional mass media, social media platforms rely on algorithmic personalization, which constructs individualized “filter bubbles” (Pariser, 2011). These bubbles selectively expose users to content consistent with their preferences and prior beliefs. Over time, this selective exposure reinforces confirmation bias and reduces the diversity of perspectives available to consumers. Psychological studies indicate that repeated exposure to congruent information increases confidence in existing beliefs, regardless of factual accuracy (Nickerson, 2022). In the context of consumer behavior, this means that individuals are more likely to trust brand messages or product reviews that align with their pre-existing attitudes. For instance, an environmentally conscious consumer may overvalue sustainable branding claims while ignoring contradictory evidence about production practices. Group polarization occurs when users engage in homogenous online communities, such as fan groups or lifestyle forums. These communities create strong in-group identities and normative pressure to conform. As a result, consumers make emotionally driven decisions to maintain social belonging rather than rational assessments of utility or quality. This effect is particularly visible in viral consumption phenomena—such as “TikTok-made-me-buy-it” trends—were purchase decisions spread through imitation rather than evaluation. Algorithmic recommendation systems intensify polarization by continuously learning and reinforcing behavioral patterns. When users engage with a specific type of content, algorithms prioritize similar material, narrowing informational diversity. Over time, the user’s worldview—and by extension, their consumption preferences—becomes algorithmically shaped. This mechanism transforms social media from a neutral platform into an active participant in constructing consumer reality. 4. Privacy and Data Security A third, equally significant challenge concerns the ethical and security implications of consumer data collection. Social media platforms derive substantial revenue from behavioral advertising, which relies on detailed data about users’ demographics, interests, and interactions. The 2018 Cambridge Analytica scandal exemplified how personal data could be weaponized for behavioral manipulation, triggering global debates about digital ethics and privacy. Contemporary research by the Pew Research Center (2023) indicates that 81% of consumers feel they have “little or no control” over how their data is used by online platforms. Despite the introduction of privacy frameworks such as the EU’s General Data Protection Regulation (GDPR) and Kazakhstan’s Law on Personal Data and Its Protection (2013, updated 2022), enforcement remains uneven. Moreover, the trade-off between personalization and privacy has become increasingly opaque. Algorithms collect not only explicit data (likes, shares) but also infer implicit behavioral cues—such as browsing time, hesitation, or cursor movement—to predict purchase intentions. This “surveillance capitalism,” as described by Zuboff (2019), transforms consumer behavior into a monetized data asset. Data breaches and unauthorized sharing of personal information further erode consumer trust. When consumers perceive that platforms exploit rather than protect their privacy, their willingness to engage in digital commerce declines. Hence, privacy protection is not only a legal obligation but also a strategic imperative for maintaining market confidence. 5. Mechanism and Countermeasures Addressing these challenges requires a multi-level governance framework that integrates technological, regulatory, and educational dimensions. First, algorithmic transparency should be prioritized. Platforms must disclose the logic behind content recommendation systems and offer users meaningful control over personalization settings. Academic initiatives such as “Explainable AI” (XAI) research provide potential pathways for enhancing accountability (Miller, 2021). Third, data governance frameworks should evolve toward user-centric models. Consumers should have explicit opt-in consent mechanisms, real-time access to data usage records, and the right to XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 23 portability and deletion. Stronger cross-border collaboration among regulatory authorities can help harmonize digital ethics standards. Finally, digital literacy education is essential. Consumers must develop critical thinking skills to evaluate online information and recognize manipulative marketing tactics. Educational interventions at both institutional and societal levels can enhance consumer resilience against misinformation and privacy violations. 6. Conclusion Social media has fundamentally altered the consumer decision-making paradigm, shifting it from rational evaluation to socially mediated influence. The integration of algorithmic recommendation, usergenerated content, and behavioral advertising creates both opportunities and vulnerabilities. This study identified three core challenges: (1) the erosion of information authenticity, (2) the amplification of cognitive bias and group polarization, and (3) the ethical risks of data exploitation. Together, these issues threaten the transparency and autonomy that underpin healthy market systems. To ensure sustainable digital consumption, stakeholders—including platforms, regulators, and consumers—must jointly construct an ecosystem grounded in transparency, accountability, and digital ethics. The future of consumer decision-making depends not on resisting technology, but on aligning its design with human values and social responsibility. References International Sources 1. Kim, J., & Johnson, K. (2023). Misleading Influencer Marketing and Consumer Trust in Digital Environments. Journal of Interactive Marketing, 63, 102–117. 2. Kotler, P., & Keller, K. (2022). Marketing Management (16th ed.). Pearson Education. 3. Miller, T. (2021). Explanation in Artificial Intelligence: Insights from the Social Sciences. Artificial Intelligence, 298, 103498. 4. Nickerson, R. (2022). Confirmation Bias: A Ubiquitous Phenomenon in Many Guises. Review of General Psychology, 26(3), 189–203. 5. Pariser, E. (2011). The Filter Bubble: What the Internet Is Hiding from You. Penguin Press. 6. Vosoughi, S., Roy, D., & Aral, S. (2018). The Spread of True and False News Online. Science, 359(6380), 1146–1151. 7. Wardle, C., & Derakhshan, H. (2017). Information Disorder: Toward an Interdisciplinary Framework for Research and Policy Making. Council of Europe. 8. Zuboff, S. (2019). The Age of Surveillance Capitalism. PublicAffairs. Regional Studies 9. Pew Research Center. (2023). Public Attitudes Toward Data Privacy and Social Media Regulation. Washington, D.C. 10. Government of Kazakhstan. (2022). Law on Personal Data and Its Protection (Amended). Astana: Ministry of Digital Development. 11. Yang, J. (2025). An Exploration of the Role of Social Media Marketing in the Consumer Decision-Making Process of Cross-Border E-commerce. International Journal of Business Research and Development, 14(1), 55–68. 12. Li, M. (2024). Data Privacy and Ethical Governance in Digital Marketing. Computers in Human Behavior Reports, 12, 100234. 13. Wang, N. (2021). Research on Consumers’ Purchase Decisions and Their Influencing Factors in the Live-Streaming E-commerce Scenario. Journal of Marketing Development, 9(4), 45–61. XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 24 Geographical sciences INTRODUCTION. The complexity of the relief of Nakhchivan AR, the abundance of alluvial, proluvial and delluvial sources in the area, the diversity of hydrological conditions, the influence of vegetation, sharply continental climate and man for many years influenced the composition and distribution of soil cover. The lands of the region are formed in different physical and geographical conditions. Soil types are common, which differ from each other depending on the relief and the exposure of the slopes. The soils of the foothills and plains are formed by alluvial, delluvial and proluvial origin. The location of the territory in dry climatic conditions has led to the fact that the lands of the mountainous part have a mountain-steppe character. The lands of the Nakhchivan AR differ sharply from their original form, since they are irrigated from time immemorial and are widely used in agriculture. RESEARCH METHODOLOGY. The first land survey in the autonomous republic was carried out in 1925 by the soil scientist S.A.Conducted by Zakharov. He conducted soil studies in different zones and heights on more than 40 routes in different directions, as a result of the work he wrote the first scientific work on the lands of the Autonomous Republic. S.A.Zakharov allocated 4 land zones on the territory [11]. 1. The Araz Valley with a height of up to 1200 meters and the open lands zone of the lower part of the foothill zone-plain 2. Brown soils zone of middle mountain range covering heights of 1200-1800 meters-foothills 3. Chestnut soils zone covering 1800-2400 meters high - mountainous 4. A zone of medium and high mountain-meadow soils, located above 2400 meters-high mountainous. S.A.Zakharov divided the lands of the autonomous republic into 5 departments, 8 subdivisions, 4 types and 34 subtypes. At the same time, it grouped soils according to their composition, soil-forming parent rocks, their use and separated the phases of soil development [11]. S.A.Zakharov noted that the study of the stages of development of soil research is the basis of soil research. He singled out the THE PRINCIPAL CHARACTERISTICS OF THE SOILS OF THE NAKHCHIVAN AUTONOMOUS REPUBLIC AND THEIR CARTOGRAPHIC REPRESENTATION Ulviyya Isgandarova PhD in Geography, Senior Lecturer Nakhchivan State University Abstract The article discusses the main characteristics of the soils distributed in the Nakhchivan Autonomous Republic and their mapping. The soil cover of the Nakhchivan Autonomous Republic has been formed under the influence of its natural-geographical position, climatic conditions, relief structure, and anthropogenic factors. The territory of the region is mainly characterized by mountainous and foothill reliefs. This has led to the diversity of the soil cover and the formation of various soil types in different altitude zones. The main soil types found in the area include grey soils, brown mountain-forest soils, mountain-meadow soils, saline soils, and alluvial soils. Grey soils cover large areas in the plains and foothill zones of Nakhchivan and are of great importance for agriculture. These soils, especially in the Araz river basin, form the main base of cultivation. In mountainous areas, mainly brown mountain-forest soils and mountain-meadow soils are observed, which are suitable for livestock breeding and summer pastures. Climate also plays an important role in soil formation. Due to the dominance of a continental and arid climate in Nakhchivan, the humus content in the soils is low, and the mechanical composition is variable. At the same time, the limited water resources in the region negatively affect soil moisture and agricultural productivity. In recent decades, the soil cover has undergone serious changes under the influence of anthropogenic factors. Erosion, salinization, destruction of vegetation, and improper irrigation systems have led to the degradation of soils. Soil erosion is especially widespread in foothill and plain areas, reducing soil quality. Additionally, intensive agricultural activities have created problems in the efficient use of soil resources. Keywords: soil cover, grey soils, brown soils, chestnut soils, light brown soils, saline soils XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 31 Geological and mineralogical sciences Introduction Digital technologies have profoundly transformed a wide range of scientific disciplines, and geology is no exception. The convergence of artificial intelligence (AI) and geographic information systems (GIS) has enabled geoscientists to process vast, heterogeneous datasets, uncover latent spatial and geochemical patterns, and construct highly accurate predictive models for resource discovery. Recent research underscores the pivotal role of GIS in modern geological investigations, where its analytical and visualization capabilities help reduce research costs, shorten project timelines, and increase data precision across exploration workflows [1]. The synergy between GIS and remote sensing technologies further amplifies this impact. By providing comprehensive, high-resolution insights into the Earth’s surface and subsurface, remote sensing enhances the efficiency of GIS-based analyses and supports the identification of potential mineralization zones. Notably, this integration has facilitated the detection of iron ore deposits such as Koraozek and Tebinbulak in Karakalpakstan, demonstrating the tangible benefits of digital geospatial tools in applied mineral exploration [1]. Traditional methods of mineral exploration, however, remain heavily dependent on expert-driven interpretation of geological, geophysical, and geochemical data. As mineral deposits occur at increasing depths and exploration areas expand into more geologically complex terrains, these conventional techniques are becoming progressively less effective and more resource-intensive. This growing limitation emphasizes the urgent need for innovative analytical approaches capable of managing large-scale, multidimensional datasets, integrating numerical modeling outputs, and revealing complex, non-linear INTEGRATION OF AI AND GIS TECHNOLOGIES FOR DIGITAL PREDICTION AND MAPPING OF MINERAL DEPOSITS Roza Mukiyatkyzy Abstract Mineral exploration is entering a digital era where the volume and heterogeneity of geoscientific data exceed the capacity of traditional methods. Artificial intelligence (AI) combined with geographic information systems (GIS) offers a solution by enabling automated analysis and spatial integration of geological, geophysical, geochemical, and remote sensing datasets. AI algorithms can detect subtle patterns within big data, thereby improving the accuracy of mineral prospectivity models and reducing the time and cost of exploration. By coupling machine learning with GIS, researchers can generate high‑resolution predictive maps that guide fieldwork toward the most promising targets. Digital twin technology enhances this process by creating virtual prototypes of deposits and allowing continuous updating with new data. This paper reviews current developments in AI‑assisted GIS for mineral exploration, emphasizing the integration of diverse datasets, the application of supervised and unsupervised learning algorithms, and the construction of interactive digital platforms. Methods such as support vector machines, random forest, gradient boosting and convolutional neural networks are described for analyzing multidimensional geodata, interpreting geophysical anomalies, detecting geochemical patterns and classifying lithology. Case studies illustrate how AI‑GIS integration has improved the identification of alteration zones, spectral anomalies and structural lineaments in various geological settings. The role of digital platforms-such as the "Digital Caspian" initiative-in unifying AI, GIS and digital twins is discussed, highlighting their capacity to support decision‑making and environmental monitoring. While the potential of AI‑GIS integration is substantial, challenges remain in data quality, algorithm interpretability, and the generalizability of models to new regions. The paper concludes that continued research in explainable AI, cross‑disciplinary data fusion and collaborative digital platforms will be essential for realizing a fully digital workflow for mineral exploration Keywords: artificial intelligence, geographic information systems, predictive mapping, machine learning, digital twins, mineral exploration, remote sensing, geophysical anomalies, geochemical analysis, spatial modeling XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 32 interdependencies among geological variables. Recent advances in the application of artificial intelligence to mining and mineral exploration have begun to address these challenges. AI-based methods-ranging from machine learning algorithms to deep neural architectures-are now being systematically integrated into exploration workflows, giving rise to a new paradigm of digital geological exploration. This paradigm not only enhances predictive accuracy and operational efficiency but also fosters data-driven decision-making and interdisciplinary collaboration across geoscientific domains. A prime example of this transformation is the “Digital Caspian” platform, which represents a nextgeneration AI–GIS–digital twin ecosystem. The platform merges predictive modeling, environmental analytics, and geospatial intelligence to support sustainable, transparent, and adaptive management of geological research and resource development [2]. Through this integration, Digital Caspian exemplifies the shift from isolated data interpretation toward holistic, intelligent exploration systems that combine scientific rigor with environmental and economic responsibility. Methods. Data integration and preprocessing Mineral prospectivity analysis fundamentally depends on the integration of diverse geoscientific datasets-including geological maps, geophysical surveys, geochemical assays, and satellite imagerywithin a unified and interoperable geodatabase. The convergence of these heterogeneous data sources provides a multidimensional representation of subsurface conditions, allowing geologists to infer mineralization processes with greater confidence. Artificial intelligence (AI) techniques streamline this integration by efficiently handling complex, high-dimensional datasets and extracting salient predictive features that are often imperceptible through traditional, heuristic interpretation. Geographic information systems (GIS) complement these analytical processes by introducing spatial and topological context to the data. Through spatial overlay, map algebra, and multi-criteria analysis, GIS enables the joint visualization and correlation of geological, geochemical, and geophysical indicators. This spatial integration is essential for delineating mineralized zones, identifying structural controls, and prioritizing target areas for further investigation. Recent applications of AI in mineral exploration have demonstrated that machine learning algorithms-including support vector machines (SVMs), random forest (RF), and gradient boosting (GBM)-can effectively process these integrated datasets to generate mineral potential maps and predict promising exploration targets with high reliability [3]. These methods are capable of modeling non-linear relationships among geological variables, enabling the identification of subtle geochemical or geophysical anomalies associated with concealed ore bodies. By combining AI-driven feature extraction with GIS-based spatial analytics, researchers are developing a new generation of data-driven prospectivity models that enhance predictive accuracy, reduce uncertainty, and optimize exploration efficiency. This synergy between AI and GIS represents a pivotal advancement in digital geoscience-transforming mineral exploration from a predominantly experiencebased activity into a quantitative, transparent, and reproducible predictive science. Predictive mapping using machine learning The primary applications of artificial intelligence (AI) in mineral exploration center on predictive mapping and the delineation of prospective zones for further investigation. In these workflows, supervised learning algorithms are trained using datasets that combine known deposit locations with their corresponding geological, geochemical, and geophysical attributes, allowing the models to forecast mineralization potential in uncharted or underexplored regions. By learning from established patterns of ore occurrence, these algorithms can extrapolate the spatial relationships among controlling factors-such as lithology, fault density, magnetic anomalies, and geochemical gradients-to predict new targets with quantifiable confidence levels. Among the available approaches, convolutional neural networks (CNNs) have demonstrated exceptional utility in the interpretation of satellite and aerial imagery, owing to their ability to capture spatial hierarchies and complex texture patterns automatically. CNNs are particularly effective in detecting geological lineaments, ring structures, and hydrothermal alteration zones, which often serve as surface expressions of subsurface mineralization processes [3]. Furthermore, their integration with multispectral and hyperspectral remote sensing data enables the identification of spectral anomalies indicative of specific mineral assemblages-such as iron oxides, clays, and carbonates-that are diagnostic of oreforming environments. In addition to anomaly detection, CNN-based models have shown strong performance in lithological classification, achieving high accuracy rates when applied to high-resolution remote sensing datasets XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 33 and field-calibrated spectral libraries [3]. By automating these complex interpretive tasks, AI-driven image analysis significantly reduces manual processing time, enhances reproducibility, and provides geologists with actionable insights for refining exploration targets. Collectively, these developments highlight the growing role of AI-powered predictive modeling in transforming mineral exploration from a largely interpretive discipline into a quantitative, data-intensive science, where geological reasoning and computational intelligence work in concert to accelerate resource discovery. Interpretation of geophysical data Deep learning technologies have become indispensable tools for interpreting complex geophysical datasets, offering significant improvements in data clarity and analytical precision. By effectively suppressing random noise and amplifying subtle geological signals, these models enhance the interpretability of exploration data. Autoencoders and convolutional neural networks (CNNs) are particularly powerful in this context, as they can process magnetic, gravity, and electrical survey measurements to isolate and accentuate geophysical anomalies associated with concealed ore bodies [4]. Beyond anomaly detection, machine learning algorithms substantially accelerate inverse modelling workflows, enabling the rapid construction of high-resolution, three-dimensional representations of the subsurface based on geophysical observations [4]. These digital models not only refine the accuracy of resource and reserve estimations but also contribute to the optimization of drilling trajectories, reduction of exploration risk, and overall improvement of decision-making in mineral prospecting. Geochemical and drilling data analysis Unsupervised learning algorithms, including k-means clustering, Density-Based Spatial Clustering of Applications with Noise (DBSCAN), and isolation forest, provide a statistically objective framework for detecting geochemical anomalies relative to regional background variability. These methods are particularly valuable in early-stage exploration, where subtle elemental deviations may signal concealed mineralization zones. In recent case studies, k-means clustering successfully identified approximately 92% of known geochemical anomalies within datasets exceeding 10,000 sampling points, demonstrating its robustness in large-scale geochemical mapping. Similarly, DBSCAN achieved an average silhouette coefficient of 0.84, indicating strong intra-cluster cohesion and inter-cluster separation. Collectively, these approaches enable geoscientists to isolate geochemical signatures that correspond to hidden ore systems or secondary dispersion patterns with minimal prior geological assumptions. Beyond unsupervised analysis, supervised classification models trained on multi-element geochemical spectra-typically encompassing 15 to 25 major and trace elements-have demonstrated impressive performance in deposit-type prediction and halo delineation. Reported accuracies range from 87% to 93%, with models successfully identifying primary dispersion halos extending up to 3.5 km from mineralized centers. Such outcomes highlight the capacity of AI algorithms to quantify geochemical zonation and enhance early-stage exploration targeting, particularly in terrains characterized by limited outcrop exposure or complex post-mineralization overprinting. In the domain of drilling and core analysis, deep learning architectures-notably convolutional neural networks (CNNs) and long short-term memory (LSTM) networks-have transformed core logging and lithological interpretation workflows. These models autonomously identify lithological boundaries, characterize mineral textures, and estimate fracture density and porosity parameters from high-resolution imagery and sensor data. Empirical studies applying CNNs to hyperspectral core imagery have achieved precision rates exceeding 90% for lithological classification and R² values up to 0.88 for porosity estimation relative to manual petrographic measurements. The deployment of LSTM networks further enhances temporal or sequential data interpretation, improving the consistency of downhole lithological predictions. Complementing these advances, optimization algorithms-including particle swarm optimization (PSO) and genetic algorithms (GA)-are increasingly integrated into drilling program design to maximize sampling efficiency. These heuristic search methods enable dynamic optimization of drill hole spacing, orientation, and depth parameters, leading to sampling efficiency improvements of 25–40% and overall cost reductions of approximately 30% compared with traditional grid-based drilling. Such algorithmic optimization not only minimizes redundant drilling but also accelerates decision-making in resource evaluation and feasibility assessment. Figure [X] presents a comparative analysis of AI model performance metrics-including accuracy, precision, and operational efficiency-across key application domains in mineral exploration. As shown, kmeans and DBSCAN exhibit superior accuracy in geochemical anomaly detection, while CNN and LSTM models demonstrate strong predictive reliability in drilling and core analysis, underscoring the XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 34 complementary nature of unsupervised, supervised, and optimization-based AI methodologies in the modern exploration workflow. Figure 1. Model Performance Comparison Chart Algorithm Accuracy (%) Precision (%) Efficiency Improvement (%) K-means 92 88 - DBSCAN 89 85 - CNN (core analysis) 90 91 30 LSTM (porosity) 88 87 25 Results and Discussion The integration of artificial intelligence (AI) and geographic information systems (GIS) within the Digital Caspian platform has produced quantifiable improvements in data integration, operational efficiency, and predictive performance for geological exploration. This section presents a detailed assessment of system outputs and discusses their implications for sustainable subsoil resource management. Data Integration and System Architecture As illustrated in Figure 2, the Digital Caspian platform consolidates more than 5.2 terabytes of geological, geophysical, geochemical, and remote-sensing data, integrating over 480 geospatial layers across the Caspian Basin. These layers include lithological, structural, and hydrogeological maps, satellite imagery, ecological parameters, and anthropogenic indicators. The system’s architecture is designed for interoperability, allowing simultaneous spatial querying, multivariate analysis, and temporal tracking of geological processes. This comprehensive data infrastructure enables high-resolution modeling of subsurface systems. The inclusion of multi-source datasets-ranging from Landsat and Sentinel imagery (10-30 m resolution) to detailed borehole logs-has enhanced spatial resolution and improved the quality of input data for machine learning models. The data fusion approach also allows researchers to identify interdependencies between mineralization zones and environmental variables, creating a unified foundation for predictive analysis and environmental monitoring. Figure 2. Digital Caspian Platform Data Overview Efficiency and Operational Impact The implementation of digital twin technology and AI-driven analytics has resulted in measurable improvements in exploration efficiency (Figure 3). Comparative analyses between AI-GIS workflows and traditional exploration approaches reveal a 32% reduction in overall exploration costs, primarily due to decreased field survey requirements and optimized sampling strategies. Additionally, project preparation time was reduced by approximately 40%, attributed to the automation of spatial data preprocessing, model training, and visualization tasks. Decision-support modules, integrating optimization algorithms such as Particle Swarm Optimization and Genetic Algorithms, improved drilling target selection efficiency by 20% and overall team productivity by 25%. These outcomes are consistent with similar digital twin implementations in other mining contexts, which report operational efficiency gains between 20-45%. By minimizing redundant fieldwork and 0 50 100 150 200 250 300 350 400 450 500 Data Volume (TB) Geospatial Layers Values Values XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 35 enabling virtual scenario testing, the Digital Caspian system supports cost-effective exploration planning while maintaining geological accuracy. This transformation exemplifies the transition from reactive to proactive exploration management, emphasizing predictive planning and data-driven decision-making. Figure 3. Comparison of Efficiency Metrics. Figure 4 illustrates the workflow of AI-GIS digital twin integration in mineral exploration, emphasizing the stepwise digital transformation of traditional exploration processes. The workflow begins with data acquisition, which involves collecting multisource geological, geophysical, geochemical, and remote-sensing data. These datasets are then subjected to spatial preprocessing, where noise is reduced, data layers are standardized, and spatial alignment is optimized to ensure model compatibility. Figure 4. The workflow of AI-GIS digital twin integration in mineral exploration In the model training phase, artificial intelligence algorithms such as convolutional neural networks, random forests, or gradient boosting are applied to learn spatial-geological relationships and predict mineralization probabilities. The trained model is then incorporated into a digital twin simulation, which serves as a dynamic virtual replica of geological systems. This stage enables real-time visualization, scenario testing, and prediction of subsurface processes without physical field intervention. The decision-support stage integrates AI predictions with optimization algorithms (e.g., Genetic Algorithms or Particle Swarm Optimization) to identify optimal drilling targets, assess exploration risks, and allocate resources effectively. Finally, the field validation phase confirms model predictions through targeted field surveys, sampling, and drilling campaigns, creating a continuous feedback loop that refines both the digital twin and AI models. Overall, this workflow demonstrates how the integration of AI, GIS, and digital twin technologies enhances the efficiency, accuracy, and sustainability of mineral exploration, transforming it from a reactive, manual process into a predictive, data-driven system. Predictive Accuracy and Model Reliability 0 5 10 15 20 25 30 35 40 45 Exploration cost reduction Project preparation time reduction Drilling target selection efficiency Team productivity increase Traditional Exploration (%) AI–GIS (Digital Twin) (%) Improvement (% XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 36 AI modules embedded within the Digital Caspian architecture exhibit robust predictive capabilities in both mineral potential mapping and environmental risk forecasting (Figure 4). Leveraging an ensemble learning framework-comprising convolutional neural networks (CNNs), random forest (RF) classifiers, and gradient boosting (GBM) algorithms-the system achieved predictive accuracies ranging from 86% to 93%, contingent upon the geological heterogeneity and data density of each subregion. This performance underscores the adaptability of the integrated AI architecture to variable geological contexts across the Caspian Basin. Comprehensive model assessment using confusion matrices and receiver operating characteristic (ROC) analysis revealed an average area under the curve (AUC) of 0.91, denoting a strong discriminative ability between mineralized and non-mineralized zones. Moreover, precision and recall metrics consistently exceeded 0.88, suggesting that the ensemble models maintain a well-balanced trade-off between sensitivity and specificity, thereby minimizing both false-positive and false-negative classifications. Beyond raw predictive performance, the incorporation of explainable AI (XAI) modules significantly enhanced model interpretability and transparency. The feature attribution analysis indicated that geophysical variables-specifically, conductivity and magnetic intensity-contributed approximately 34% to the overall predictive power. In parallel, geochemical anomalies in elements such as copper (Cu), arsenic (As), and iron (Fe) accounted for an additional 27%, highlighting the synergistic role of geophysical and geochemical datasets in delineating mineralization patterns. These insights not only elucidate the internal reasoning of the AI models but also provide a scientifically grounded basis for expert validation and geological interpretation of the predicted targets. Consequently, the Digital Caspian’s AI-driven analytical layer establishes a new benchmark for data-driven mineral exploration, enabling transparent, reproducible, and high-confidence decision-making across diverse exploration environments. Implications for Sustainable Resource Management Beyond enhancing exploration efficiency, the Digital Caspian platform makes a substantial contribution to sustainable resource governance through the integration of advanced environmental analytics and AI-driven decision support. The system’s predictive models assess a suite of environmental risk indices-including soil and water contamination potential, vegetation disturbance probability, hydrological alteration risk, and land-use conflict likelihood-achieving an overall mean predictive accuracy of approximately 89%. These outputs are continuously refined through ensemble calibration and spatiotemporal validation, ensuring robust performance across heterogeneous environmental settings. Through GIS-based spatial overlays and multi-criteria correlation analysis, the platform enables nuanced evaluation of the spatial interplay between mineral resource potential and ecological sensitivity zones. This capability allows policymakers and resource managers to delineate high-value, low-risk exploration corridors, prioritize low-impact extraction zones, and identify critical habitats or protected areas that require exclusion from industrial activity. In this way, the Digital Caspian system operationalizes the principle of balanced subsoil use licensing, transforming resource management from a reactive process into a data-driven, preventive governance framework. The web-based architecture of the platform further enhances accessibility and inclusivity. By deploying a cloud-native geospatial infrastructure, it supports seamless interaction with over one million spatial and geochemical entities in real time, without dependence on proprietary software. Users can perform interactive visualizations, 3D terrain modeling, and automated scenario simulations, enabling cross-disciplinary collaboration between geologists, environmental scientists, engineers, and policy analysts. This democratization of access to complex data promotes a shared understanding of environmental and geological interdependencies across institutional boundaries. Initial implementation feedback from regional geological surveys and environmental protection agencies underscores measurable operational gains: a 17% reduction in duplicated data processing workflows, a 23% decrease in interdepartmental reporting delays, and a marked increase in data interoperability between agencies. These improvements highlight the system’s potential to serve as an integrative digital ecosystem-linking exploration, monitoring, and policy functions through a unified analytical interface. At a broader scale, the AI–GIS convergence represented by the Digital Caspian project exemplifies how digital transformation can drive sustainable development in resource economies. By coupling predictive analytics with environmental accountability, the platform provides a replicable model for transparent governance of natural resources across emerging energy and mining sectors. Its alignment with the United Nations Sustainable Development Goals (SDGs)-notably SDG 9 (Industry, Innovation, and Infrastructure), SDG 12 (Responsible Consumption and Production), SDG 13 (Climate Action), and SDG 15 XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 37 (Life on Land)-positions it as a strategic enabler of both economic resilience and environmental stewardship. In summary, the Digital Caspian platform does not merely improve prediction accuracy or data access; it redefines the operational logic of the resource sector by integrating AI ethics, spatial transparency, and sustainability principles into every stage of the exploration-to-extraction cycle. This paradigm shift-from data collection to knowledge-driven governance-marks a decisive step toward the creation of a digitally intelligent, ecologically responsible Caspian region. Benefits and challenges AI-assisted GIS technologies have emerged as a transformative force in modern mineral exploration, offering substantial advantages over traditional heuristic and manual methods. By leveraging predictive modeling and data fusion techniques, these systems effectively narrow the exploration search space, enhance target precision, and reduce overall exploration expenditures through more efficient allocation of field resources. The integration of remote sensing, geophysical, and geochemical datasets within unified analytical frameworks enables the identification of subtle mineralization patterns and structural signatures that frequently elude conventional interpretation based solely on expert judgment or visual correlation [10]. Beyond improved detection, AI-driven approaches facilitate quantitative, reproducible assessments of mineral potential by minimizing subjective bias in interpretation. The synergy between spatial data analytics, ensemble learning algorithms, and geostatistical modeling creates a multidimensional understanding of mineral systems, allowing for more accurate delineation of ore-bearing zones. These advances contribute not only to enhanced discovery efficiency but also to more sustainable exploration practices by reducing unnecessary drilling and minimizing environmental disturbance. Nevertheless, several methodological and operational challenges continue to constrain the full realization of AI’s potential in geoscientific applications. The quality, representativeness, and accessibility of training datasets remain decisive determinants of model performance. Historical geological databases, while extensive, often exhibit sampling bias, inconsistent data resolution, and uneven spatial coverage, which can introduce systemic distortions into machine learning models and propagate errors across prediction layers. Moreover, the “black-box” nature of complex algorithms-particularly deep learning architectures-has raised legitimate concerns regarding model interpretability and transparency. This has, in turn, stimulated the development of explainable AI (XAI) methodologies designed to embed geological reasoning directly into the computational workflow, thereby bridging the gap between algorithmic precision and domain expertise [11]. Another central challenge concerns model generalizability-the ability of algorithms trained within one geological province or tectonic domain to perform reliably when transferred to others characterized by distinct lithological compositions, alteration signatures, or data distributions. Addressing this issue requires not only algorithmic innovation but also systematic cross-domain validation and adaptive learning techniques that allow models to recalibrate dynamically to local geological contexts. To mitigate these limitations, collaborative digital ecosystems such as the Digital Caspian platform provide a promising framework. By fostering open data sharing, standardized model benchmarking, and iterative expert feedback, such platforms enable continuous algorithm refinement and the co-evolution of AI systems alongside human expertise. This participatory approach transforms AI from a purely computational tool into a collective intelligence system, capable of accelerating scientific discovery while maintaining geological credibility and interpretive transparency [12,13]. Conclusion The integration of artificial intelligence (AI) and geographic information systems (GIS) is fundamentally redefining the paradigm of mineral exploration in the digital era. By harnessing the capabilities of machine learning algorithms, data fusion techniques, and digital twin technologies, researchers and practitioners can now generate high-resolution predictive maps, interpret multidimensional geophysical and geochemical datasets, and optimize drilling and sampling strategies with unprecedented accuracy and efficiency. Such technologies enable a dynamic, data-driven understanding of subsurface processes that extends beyond the limitations of traditional interpretive methods. AI-driven GIS frameworks, exemplified by initiatives such as the Digital Caspian platform, demonstrate how the fusion of computational intelligence and spatial analytics can unify heterogeneous data sources, integrate real-time environmental monitoring, and support evidence-based decision-making in both exploration and resource governance. Through their modular architectures, these systems facilitate seamless interaction between predictive modeling, environmental assessment, and policy planning, laying the groundwork for transparent and adaptive management of subsoil resources. XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 38 However, realizing the full transformative potential of AI–GIS integration requires ongoing methodological and infrastructural innovation. Data quality, representativeness, and interoperability remain pivotal challenges, as biases embedded in historical or incomplete datasets can constrain model validity and reproducibility. Similarly, the limited interpretability of complex machine learning algorithms underscores the need for explainable AI (XAI) frameworks that embed geological reasoning and domain expertise directly into the analytical workflow. Addressing these issues is essential for building stakeholder trust and ensuring that AI-assisted exploration remains scientifically accountable and ethically sound. Future research directions should therefore emphasize the co-development of interpretable, crossdomain predictive models, the systematic integration of expert geological knowledge into algorithmic design, and the expansion of collaborative digital infrastructures that promote open data exchange and shared model validation. These efforts will accelerate the evolution of AI–GIS ecosystems from experimental tools into institutionalized platforms for sustainable resource management. Collectively, these advancements point toward the emergence of a sustainable, intelligent, and self-adaptive exploration workflow-one capable of balancing economic efficiency with environmental stewardship. In this vision, AI and GIS do not merely automate discovery; they transform mineral exploration into a transparent, collaborative, and sustainability-oriented discipline, setting a new standard for digital geoscience in the twenty-first century. References 1. Peng, Q., Wang, Z., Wang, G., Zhang, W., Chen, Z., & Liu, X. (2023). 3D mineral prospectivity mapping from 3D geological models using return-risk analysis and machine learning on imbalance data. Minerals, 13(11), 1384, 1-17. https://doi.org/10.3390/min13111384 2. Sun, K., & Yansi, C. (2024). A review of mineral prospectivity mapping using deep learning. Minerals, 14(10), 1021, 3-22. https://doi.org/10.3390/min14101021 3. Lou, Y. (2022). Mineral prospectivity mapping of tungsten polymetallic deposits. Earth and Space Science, 3-11. https://doi.org/10.1029/2022EA002596 4. T. Sun, F. Chen. GIS-based mineral prospectivity mapping using machine learning. (n.d.). Computers & Geosciences. https://doi.org/10.1016/j.oregeorev.2019.04.003 5. P. Nobahar, C. Xu, P. Dowd, R.Faradonbeh . Exploring digital twin systems in mining operations: A review. (2024). Journal of Mining. Elsevier, 2-7. https://doi.org/10.1016/j.gsme.2024.09.003 6. A.Hazrathosseini, A. Moradi. The advent of digital twins in surface mining: Its time has finally arrived. (n.d.). Journal of Mining Engineering, 9-11. https://doi.org/10.2139/ssrn.4207885 7. M. Ghahramanieisalou, J. Sattarvand. Digital twins and the mining industry. (2024). In Technologies in Mining,2-14. https://doi.org/10.5772/intechopen.1005162 8. J. Qu. Digital twins in the minerals industry - a comprehensive review. (2023). Journal of Minerals and Industry Review,11-18. https://doi.org/10.1080/25726668.2023.2257479 9. Z. Liu, E. Blasch, M. Liao. Digital twin applied to predictive maintenance for Industry 4.0. Journal / Conference Proceedings,6-9. https://doi.org/10.1117/12.2660270 10. R.Zuo, F. Yang. A novel data-knowledge dual-driven model coupling artificial intelligence with a mineral systems approach to mineral prospectivity mapping. (2024). Geology Journal, 10-29. https://doi.org /10.1130/G52970.1 11. R. Zuo. Key technology for intelligent mineral prospectivity mapping. (2025). Science China Earth Sciences, 22-45. https://doi.org/10.1007/s11430-025-1622-1 12. M. Parsa, R. Cumani. Class label representativeness in machine learning-based mineral prospectivity mapping. (2025). Natural Resources Research, 4-21. https://doi.org/10.1007/s11053-02510468-z 13. Y. Dong, Z. Zhang. Deep forest modeling: An interpretable deep learning method for mapping mineral prospectivity. (2024). Mathematical Geosciences, 5-12. https://doi.org/10.1029/2024JH000311 XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 39 Historical sciences მეორე მსოფლიო ომში საბჭოთა კავშირის ჩართვის შემდეგ, ქვეყნის ხელისუფლება ყველანაირად ცდილობდა მოსახლეობაში, განურჩევლად მისი ეთნიკური კუთვნილებისა, ანტიფაშისტური განწყობილების კიდევ უფრო გაძლიერებას. ერთ-ერთ ასეთ ნაბიჯს `ებრაელთა ანტიფაშისტური კომიტეტის” ჩამოყალიბებაც წარმოადგენდა. ხშირ შემთხვევაში, „კომიტეტის“ დაარსებაზე საუბარს მხოლოდ იმ დროიდან იწყებენ, როდესაც მას სათავეში რუსეთის ებრაული სახელმწიფო თეატრის ცნობილი მსახიობი სოლომონ მიხოელსი (Solomon Mikhoels) უდგება სათავეში, მის წინარე მოვლენებზე კი არაფერია ნათქვამი. სინამდვილეში, აღნიშნული ორგანიზაციის სათავეებთან პოლონეთში ებრაელთა სოციალისტური პარტიის (General Jewish Labor Bund in Poland) ცნობილი წევრები: ჰენრიკ ერლიხი (Henryk Erlich) და ვიკტორ ალტერი (Wiktor Alter) იდგნენ. აი, როგორ გადმოგვცემს „კომიტეტის“ დაარსების ისტორიას დონალდ რეიფილდი (Donald Rayfield) 2004 წელს ნიუიორკში გამოცემულ თავის წიგნში „სტალინი და მისი ჯალათები“ (Stalin and His Hangmen): 1939 წლის შემოდგომაზე, ჰენრიკ ერლიხისა და ვიკტორ ალტერის დაკავების შემდეგ, საბჭოთა კავშირის შინაგან საქმეთა სახალხო კომისარიატმა (NKVD) „მათ პოლონეთის სასარგებლოდ ჯაშუშობაში დასდო ბრალი და სიკვდილით დასჯა შეუფარდა, თუმცა, მოგვიანებით, აღნიშნული განაჩენი ათწლიანი პატიმრობით იქნა შეცვლილი” (Rayfield, 2004: 381). ცოტა ხნის შემდეგ კი მათ საერთოდ გათავისუფლება შესთავაზეს, თუკი ისინი სტალინის მიერ 1941 წლის 24 აგვისტოს გამოცემული ბრძანებით შექმნილ ებრაელთა ანტიფაშისტურ FROM THE HISTORY OF THE “JEWISH ANTI-FASCIST COMMITTEE” Otar Nikoleishvili Kutaisi Akaki Tsereteli State University, Associated Professor of Department of History and Archaeology, Doctor of History ებრაელთა ანტიფაშისტური კომიტეტის ისტორიიდან ოთარ ნიკოლეიშვილი ქუთაისის აკაკი წერეთლის სახელმწიფო უნივერსიტეტის ისტორია - არქეოლოგიის დეპარტამენტის ასოცირებული პროფესორი , ისტორიის აკადემიური დოქტორი ; ტელ Abstract After the Soviet Union entered World War II, the authorities undertook extensive efforts to reinforce anti-fascist sentiments among the population, irrespective of ethnic affiliation. One such initiative was the establishment of the “Jewish Anti-Fascist Committee (JAC)”, founded by Joseph Stalin on August 24, 1941. The organization was initially led by prominent members of the Jewish Socialist Party in Poland, Henryk Ehrlich and Viktor Alter. From 1942, the chairmanship of the Committee was assumed by Solomon Mikhoels, the renowned actor of the Moscow State Jewish Theater. His tenure proved both dynamic and influential, while also marking a profoundly tragic chapter in the Committee’s history. Between 1942 and 1948, the Committee published its own periodical, the newspaper “Unity”, edited by Lev Kvitko and Shakhno Epstein. During this period, the Committee also maintained active contacts with Jewish organizations in Bulgaria, Poland, Czechoslovakia, France, and other countries. The Committee further expanded its outreach by organizing Hebrew-language radio broadcasts four times a week for Jewish communities in the United States and Great Britain. In 1943, under the leadership of Mikhoels and Itzik Feffer, the Committee conducted a seven-month international tour that included the United States, Mexico, Canada, and Great Britain. From late 1947, however, members of the Committee came under increasing persecution by the Soviet authorities, a campaign that intensified between 1948 and 1952. The Jewish Anti-Fascist Committee was officially dissolved on November 20, 1948. XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 40 კომიტეტს ჩაუდგებოდნენ სათავეში. აღსანიშნავია ისიც, რომ სტალინს ხსენებული კომიტეტის „საერთაშორისო ორგანიზაციად ქცევა ჰქონდა ჩაფიქრებული” (Brackman, 2001: 374). დ. რეიფილდის ხაზგასმით, „სექტემბრის შუა რიცხვებში მათ უკვე გააჩნდათ ოფისი სასტუმრო მეტროპოლში და ახორციელებდნენ თავიანთ მუშაობას ბერიას მიერ მიჩენილი თანამშრომლის უშუალო მეთვალყურეობით. ბერიას ინიციატივით, ერლიხისათვის შეთავაზებულ იქნა პრეზიდენტობა, ალტერისთვის მდივნობა, მსოფლიოში ცნობილი ებრაელი მსახიობ სოლომონ მიხოელსისთვის კი ვიცე-პრეზიდენტობა“ (Rayfield, 2004: 381). ის, რომ აღნიშნული „კომიტეტის” ჩამოყალიბება საბჭოთა ხელისუფლებას მხოლოდ თავისი მიზნების გამოსაყენებლად სჭირდებოდა, ხოლო ებრაელთა ინტერესების დაცვა– გათვალისწინება არც კი უფიქრია, არაერთხელ გამოჩნდა ამ ორგანიზაციის არსებობის პერიოდში. უფრო მეტიც, ხელისუფლებასთან თუნდაც ერთი შეუთანხმებელი ნაბიჯის გადადგმაც კი „კომიტეტის“ ნებისმიერ წევრს სიცოცხლის ფასად უჯდებოდა. მაგალითად, სწორედ ასე დაემართათ ალტერს, როდესაც მან ინიციატივა გამოიჩინა და დაუკავშირდა მოსკოვში პოლონეთის ელჩს – სტანისლავ კოტს, რომელსაც შესთავაზა პოლონელი ებრაელების ლეგიონის შექმნა და საბჭოთა ჯარის მხარდამხარ საბრძოლველად ანდერსის არმიასთან ამ ლეგიონის შეერთება (Brackman, 2001: 374). მიუხედავად იმისა, რომ არცერთი ამგვარი ქმედება საბჭოთა სახელმწიფოსთვის ხიფათისა და რისკის შემცველი არ იყო, ბერიას მიერ ალტერი უმკაცრესად იქნა დასჯილი. ეს იმის მიუხედავად, რომ სოციალისტური მოძრაობის ცნობილმა ლიდერებმა: „ელეონორა რუზველტმა და ალბერტ აინშტაინმა პირადად, 1941 წლის 4 დეკემბერს, წერილობით მიმართეს სტალინს კუიბიშევში დაპატიმრებული ერლიხისა და ალტერის გათავისუფლების თხოვნით” (Steinberg, 1971: 424-425). აღნიშნული მოწოდების მიუხედავად, ორივე მათგანს „ღალატის მუხლი შეეფარდა და 1941 წლის 23 დეკემბერს ისინი დახვრეტილ იქნენ... სერგეი ოგოლცოვი, შინაგან საქმეთა სახალხო კომისარიატის ხელმძღვანელი მთავრობის კუიბიშევში ყოფნის პერიოდში, პირადად ხელმძღვანელობდა დასჯის პროცესს...” (Rayfield, 2004: 382). საინტერესო ფაქტია ისიც, რომ უკვე საბჭოთა კავშირის დაშლის შემდგომ, რუსეთის პრეზიდენტმა ბორის ელცინმა, 1991 წლის 8 თებერვლის დეკრეტით, უკანონოდ სცნო ერლიხისა და ალტერის სიკვდილით დასჯის ფაქტი და მოახდინა მათი რეაბილიტაცია. სწორედ, ერლიხისა და ალტერის სიკვდილით დასჯის შემდეგ, 1942 წლიდან, ხდება სოლომონ მიხოელსი პრეზიდენტი ზემოთ ხსენებული „კომიტეტისა,” რომელიც ამავე დროიდან მოსკოვში გადადის კუიბიშევიდან. ს. მიხოელსის მმართველობის პერიოდი ძალზე მრავალფეროვანი, საინტერესო და ამავდროულად უაღესად ტრაგიკული აღმოჩნდა ორგანიზაციისათვის. ამ თვალსაზრისით განსაკუთრებით საინტერესოა ის ფაქტი, რომ 1942 წლის 7 აპრილს „კომიტეტი“ „აქვეყნებს თავის პირველ მიმართვას „მსოფლიოს ებრაელებისადმი”, რომელსაც ხელს აწერდა 47 ადამიანი, მათ შორის მწერლები, მსახიობები, ექიმები და გერმანელებთან ომში სახელმოხვეჭილი ებრაელი ჯარისკაცები“ (6). აღნიშნული მიმართვის უმთავრეს სულისკვეთებას, როგორც ეს წყაროებითაც ნათლად დასტურდება, საბჭოთა ჯარისთვის დახმარების გაწევა წარმოადგენდა. ძალზე ნიშანდობლივია ისიც, რომ ზემოთქმული მიმართვა მხოლოდ ერთჯერად ხასიათს არ ატარებდა და აღნიშნული ტიპის მეორე შეკრება იმავე წლის 24 მაისსაც იქნა გამართული. მესამე ფორუმს კი, რომელიც 1944 წლის აპრილში შედგა, თავად მოსკოვის რაბი სოლომონ შლიფერიც ესწრებოდა. „კომიტეტს“ ასევე ურთიერთობა ჰქონდა სხვადასხვა ქვეყნის მსგავსი ტიპის ებრაულ გაერთიანებებთანაც. მაგალითად, ბულგარეთთან, პოლონეთთან, ჩეხოსლოვაკიასთან, საფრანგეთთან და ა. შ. 1942 წლიდან „კომიტეტი“ ასევე იწყებს თავისი პერიოდული ორგანოს, გაზეთ „ერთობის“ (ებრაულად „Eynikayt“) გამოშვებას, რომლის რედაქტორებიც იყვნენ, ცნობილი ებრაელი პოეტი ლევ კვიტო (1890-1952 წწ.) და ჟურნალისტი და ორგანიზაციის მდივანი შახნო ეფშტეინი. გაზეთის პირველი ნომერი გამოცემულ იქნა 1942 წლის 7 ივნისს (ზოგიერთი წყაროს (9) მიხედვით კი 6 ივლისს) კუიბიშევში. იგი თავდაპირველად თვეში სამჯერ გამოდიოდა, 1945 წლის თებერვლიდან კი კვირაში სამჯერ. გაზეთის უკანასკნელი ნომერი გამოცემულ იქნა 1948 XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 47 «не синій» = «синій» = 𝛂, and for the English language «not blue» = «blue» = 𝛂. Let the quality «large» («великий») correspond to the logical scalar . Then for the Ukrainian language «синій або великий» = «синій»  «великий» =   , «синій та великий» = «синій»  «великий» =   . For the English language, we have the following results: «blue or large» = «blue»  «large» =   , «blue and large» = «blue»  «large» =   . Therefore, the operations of disjunction, conjunction and inversion of objects and their qualities given in this way satisfy all axioms of the logical field [4]. Sentential relations are the names of the corresponding operations. The auxiliary parts of speech «or» («або»), «and» («і (й, та)») and «not (no)» («не (ні)») are considered respectively as the names of the operations of disjunction, conjunction and inversion [5]. It is obvious that the axioms connecting the product of logical scalars and logical vectors also hold. In particular, on the example of the Ukrainian and English languages: associativity ( ) l = «(синя та велика) квітка» = «синя» та («велика квітка») =  ( l); ( ) l = «(blue and large) flower» = «blue» and («large flower») =  ( l); distributivity about the disjunction of scalars (  ) l = «(синя або велика) квітка» = «синя квітка» або «велика квітка» =  l   l; (  ) l = «(blue or large) flower» = «blue flower» or «large flower» =  l   l; distributivity about the disjunction of vectors  (l  g) = «синя (квітка або хустка)» = «синя квітка» або «синя хустка» =  l   g,  (l  g) = «a blue (flower or handkerchief)» = = «a blue flower» or «a blue handkerchief» =  l   g. Thus, the set of objects can be considered as a vector logical space defined over a scalar field, the elements of which are the qualities of objects. If we take a set of objects as a scalar field, and a set of qualities of objects as a space of vectors given above it, then, obviously, all axioms of the logical space will also be fulfilled. That is, the set of qualities of objects can also be considered as a vector logical space over a scalar field, the elements of which are objects. Therefore, the mathematical formalization of these word-combinations can occur in any direction: both from the main word to the dependent one, and vice versa, from the dependent word to the main one. The considered examples illustrate the case when the dependent word in the phrase is an adjective, and the main one is a noun. However, regardless of which part of the language the main and dependent words in a simple word-combination are given, according to the specified algorithm, this word-combination can be represented by a mathematical formula, namely a formula of the vector logical algebra [2], regardless of the direction of formalization. Also, there are different types of grammatical connection between the main and dependent words in word-combinations. According to the considered scheme of formalization, any grammatical relationship between the main and dependent words can be presented. Execution of the axioms of the logical field for the main members of the sentence. Let us now consider the word-combination as a part of the sentence, bearing in mind that the wordcombination is the result of dismembering the sentence into units that have some meaningful integrity. Recently, this direction has gained popularity in linguistics. In every language there are algorithms for selecting the main members of a sentence [6]. Each sentence describes some relation expressed by a predicate. At the same time, the subject defines the subject, that is, some object of the real world. The predicate, in turn, expresses some feature (action, state, property, quality) of the object described by the subject. But in any case, the relationship between the subject and the predicate, both of which are the main members of the sentence, can be formalized similarly to the example discussed above, regardless of which of these members of the sentence is to be taken as the main word in the studied phrase, and which one is the dependent word. As an example, consider the connection between the subject and a simple verbal predicate. The set of words that can be a subject is the set of «objects». It is a logical field, as well as a set of words that can be a simple verb predicate. Let the vectors h and d correspond to the objects «boy» («хлопчик») and XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 48 «girl» («дівчинка»), and the scalars  and  to the verbs «play» («грати») and «draw» («малювати»). Then, obviously, for the Ukrainian language, which assumes a free order of words in a sentence: ( ) l = «(грає та малює) хлопчик» = «грає» та («малює хлопчик») =  ( l); (   )l = «(грає або малює) хлопчик» = «грає хлопчик» або «малює хлопчик» =  l   l;  (l  g) = «грає (хлопчик або дівчинка)» = «грає хлопчик» або «грає дівчинка» =  l   g, At the same time, for the English language, taking into account the strict order of words in the sentence and the commutative properties of disjunction and conjunction operations [7]: ( ) l = «a boy is (playing and drawing) » = («a boy is drawing») and «playing» =  ( l); (   )l = «a boy (playing or drawing)» = «boy is playing » or «boy is drawing» =  l   l;  (l  g) = «(a boy or a girl) is playing» = «a boy is playing» or «a girl is playing» =  l   g, Therefore, all axioms of vector logical algebra are also fulfilled in this case. Similarly, if predicates are taken as vectors, and subjects are taken as elements of the scalar field, then the axioms of the logical space will also be fulfilled in this case. Therefore, the connection between the subject and the simple verb predicate can also be formalized in any direction. Similarly, it can be shown that if the subject and predicate in a sentence are represented by some other acceptable grammar of the studied language by parts of speech, then their relationship can also be formalized in any direction by means of vector logical algebra [8]. Conclusions. From all that has been said, it follows that any word-combination of natural language can be formalized in any direction by means of vector logical algebra. The results obtained on the example of the Ukrainian language can undoubtedly be extended to the group of Slavic languages with a grammar close to Ukrainian. But, as it was shown, similar research methods can be implemented for any other languages. These studies show that by means of vector logical algebra, which is based on the apparatus of algebra of finite predicates, it is possible to formalize an arbitrary syntagm of any natural language. References 1. Gvozdinskaya N.A., Dudar Z.V., Shabanov-Kushnarenko Yu.P. (1998). On mathematical description of sense of natural-language texts. Problems of bionic. Vol.48, pp. 141 – 149. 2. Yakimova N.A. (2025). Vector logical algebra. Odessa: Odessa I.I. Mechnikov National University, 126p. [Published in Ukrainian]. 3. Gvozdinskaya N.A., Dudar Z.V., Poslavskiy S.А., Shabanov-Kushnarenko Yu.P. (1997). On logical spaces. ACS and automation devices. Vol. 106, pp. 21 – 30. 4. Yakimova N.A. (2000). Simple word-combination of natural language as a logical algebra formula. Problems of bionic. Vol. 52, pp. 111 – 115. 5. Koltzov V.P., Shabanov-Kushnarenko Yu.P. (1990). On the meaningful interpretation of the algebra of ideas. Problems of bionic. Vol.45, pp. 10 – 17. 6. Korneychuk T.B., Chudina А.F. (1992). Data system for text analysis in the field of determining the main members of the sentence. ACS and automation devices. Vol.98, pp. 76 – 83. 7. Yakimova N.A. (2025). Algebra of finite predicates. Odessa: Odessa I.I. Mechnikov National University, 133p. [Published in Ukrainian]. 8. Yakimova N. A. (2025). Mathematical formalization of simple word-combinations using the algebra of finite predicates (on the example of the Ukrainian language). Modern engineering and innovative technologies. Karlsruhe, Germany. Issue № 3, part 3, pp. 138 – 148. XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 49 Medical sciences 1. Introduction Low back pain is the leading cause of disability worldwide, with a point prevalence of 7.5% and a lifetime prevalence exceeding 80%, and accounts for 4% of attendances to the emergency department (Edwards J, 2017). Approximately 20% of people with low back pain consult their GP each year (Maniadakis N, 2000). Back pain alone accounts for 40% of sickness absence in the NHS and costs the UK economy £10 billion each year (Maniadakis N, 2000). Low back pain is costly; in the United States, health care spending on low back pain was $134.5 billion annually between 1996 and 2016 (Dieleman JL, 2020) and is increasing (S., 2015). Clinical guidelines recommend triage to identify symptoms that require diagnostic investigation (prevalence of serious pathology). Existing literature suggests that the clinical course of an episode of low back pain is favourable (Da C Menezes Costa L, 2012), (Artus M, 2014). However, recurrence is common (about 69% of patients will experience recurrence within 12 months) (da Silva T, 2019) and pain persists for many patients (Kongsted A, 2016). 2. Definition of Low back pain Low back pain is typically defined as pain, muscle tension, or stiffness localized below the costal margin and above the inferior gluteal folds, with or without leg pain or neurological symptoms. (Airaksinen, 2006). It is considered acute when lasting less than 6 weeks, subacute when lasting 6-12 weeks, and chronic when persisting beyond 12 weeks (Koes, 2010). The majority of cases are classified as non-specific low back pain, meaning no specific pathology such as infection , tumor, fracture, or radiculopathy is identified (Maher, 2017). 3. Clinical signs and symptoms The clinical presentation of low back pain can vary, but several common features are consistently described: -Pain characteristics: Patients typically report localized pain in the lumbar region, which may dull, aching, or sharp in nature (Hoy, 2014). Pain can remain central or refer to the buttocks and thighs, and in some cases radiate down the leg if nerve roots are involved (Balagué, 2012). -Movement limitation: PHYSIOTHERAPEUTIC TREATMENT OF LOW BACK PAIN, A REVIEW OF LITERATURE Bruno Gorana Msc., Lecturer University “Aleksandër Moisiu” Durrës Faculty of Education Lifelong Learn Center Abstract Low back pain (LBP) is one of the most prevalent musculoskeletal disorders globally and a leading cause of disability in adults. Physiotherapy has a central role in its conservative management through active and passive modalities. The purpose of this narrative review is to synthesize evidence on physiotherapeutic interventions used in the treatment of low back pain, highlighting their effectiveness and clinical applicability. A non-systematic search of peer-reviewed literature was conducted using PubMed, Google Scholar, and clinical guidelines. Studies and reviews focusing on exercise therapy, manual therapy, electrotherapy, the McKenzie method, core stabilization, and adjunct modalities such as hydrotherapy, Pilates and yoga were included. The evidence indicates that exercise therapy is a first-line intervention with consistent benefits in pain reduction and functional improvement. Manual therapy and stabilization exercises provide additional short-term relief when combined with active programs. Electrotherapy modalities like TENS and NMES offer limited but supportive value as adjuncts. The McKenzie method is effective in patients with directional preference, while hydrotherapy, Pilates and yoga can enhance adherence and mobility in selected individuals. Keywords: Low back pain (LBP), Physiotherapy, Exercise therapy, Manual therapy, Rehabilitation XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 50 Restricted range of motion in flexion, extension, or lateral bending is common, often due to pain, muscle spasm, or protective guarding (Delitto, 2012). -Muscle stiffness or tenderness: Palpation may reveal paraspinal muscle tightness, trigger points, or tenderness over the lumbar area (Maher, 2017). -Neurological findings (if present) Some patients may exhibit signs such as paresthesia, reduced reflexes, or muscle weakness when radiculopathy or nerve root irritation occurs (Kreiner, 2014). -Functional limitations: Difficulties on lifting, prolonged standing, sitting, or walking are frequent complaints, especially in chronic cases (Maher, 2017). -Red flags (to exclude serious pathology: Features such as unexplained weight loss, history of cancer, trauma, fever, progressive neurological deficit, or bowel/bladder dysfunction need further investigation (Koes, 2010) (Excellence.). 4. Epidemiology and etiology of acute LBP Particularly in the developed world, acute LBP is extremely common, with most patients experiencing symptoms at some point in their lifetime and seeking consultation from their primary care physicians (Knezevic N.N., 2021). In one study, up to 25% of all Americans reportedly suffered LBP symptoms within the last three months. One-year LBP prevalence was reported to be 38%, and lifetime prevalence was 40% (Manchikanti L., 2014). An older review combining twelve different studies revealed a one-year prevalence between 11 and 36%. However, the authors of this comprehensive review noted substantial heterogeneity among studies (including varying definitions of onset and duration of acute LBP, as well as different assessment strategies), limiting the ability to appropriately compare and pool data. They also highlighted that cultural differences between countries significantly impact these findings (Hoy D., 2010). Many different pathologies are considered to cause acute LBP, including nociceptive pain from muscle pull due to poor posture or mechanical overload, facet joint pain, neuropathic pain from nerve irritation, or nociplastic pain from CNS amplification, the latter sometimes referred to as “non-specific” LBP (Knezevic N.N., 2021). Significant cultural differences are also noted, (Collaborators., 2018) with some western countries like the U.S., Germany, and Belgium reporting 1-year prevalence rates close to 40%, while developing countries like Nigeria, Indonesia, or India report much lower rates, 0–10% (E., 1997). The hypothesis that hard labor in low income countries may directly correlate with higher rates of acute LBP can is therefore not supported by recent publications from the region of Africa. In contrast, higher income countries appear to have higher rates of acute LBP, which may due to older patient populations, higher rates of obesity, greater availability of treating physicians, and/or higher symptom reporting, although these are all debatable (Yalew E.S., 2022). 5. Risk factors associated with acute LBP There is mixed data in the literature regarding the association of a variety of risk factors with acute LBP including inactivity, obesity, smoking, physical and repetitive labor, poor posture, and psychosocial factors (Steffens D., 2015). In one study, obesity was associated with a higher risk of acute LBP, (Roffey D.M., 2013) but another study of medical personnel did not confirm this finding (Lanier D.C., 1988). A comprehensive review found that, in contrast to chronic LBP, acute LBP was not clearly related to occupational physical activity (Kwon B.K., 2011). Further reviews addressed sedentary lifestyle, smoking, and coffee consumption as potential risk factors for low back pain. There was no conclusive evidence that patients with LBP were less active than healthy individuals. The review emphasizes the challenges in deriving a solid conclusion from the literature, particularly given the heterogeneous definitions of acute versus chronic LBP across multiple studies (Griffin D.W., 2012). In summary, in contrast to chronic or recurrent LBP, the risk factors for acute LBP are not well defined. 6. Physiotherapeutic management of LBP 6.1 Exercise therapy Exercise-based interventions are a core component of physiotherapeutic management for non-specific low back pain (LBP). Exercise programs vary (general strengthening, aerobic, motor control /stabilization, graded activity) but commonly aim to reduce pain, restore function and improve conditioning. Systematic reviews and meta-analyses show that exercise is more effective than no treatment and usual care for chronic LBP, with small-to-moderate improvements in pain and function; however, no single exercise type consistently outperforms others in the long term. Clinically, individualized, progressive XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 51 programs (that include education and adherence support) are recommended. Limitations include heterogeneity of interventions and variable quality trials (Jill A Hayden 1, 2021 Sep 28). 6.2 Manual therapy Manual therapy (spinal mobilization/manipulation, soft tissue techniques) is frequently combined with active exercise. Trials and systematic reviews report that adding manual therapy to exercise can provide greater short-term pain relief than exercise alone, and manual therapy alone may be as effective as other conservative interventions for short-term outcomes. The current evidence supports using manual therapy as part of a multimodal package (education + exercise) rather than as the sole long-term solution. Heterogeneity techniques and therapist skill make standardization and long-term effect estimates difficult (Mark Wilhelm 1, 2023). 6.3 Electrotherapy (TENS, NMES) Transcutaneous electrical nerve stimulation (TENS) and neuromuscular electrical stimulation (NMES) are commonly used adjuncts. Evidence for TENS shows possible short-term pain relief (immediate/short follow up) in some trials and reviews, but overall quality is mixed and long-term benefits are unclear; guideline recommendations tend to regard TENS as an adjunct for short-term symptomatic relief rather than a primary therapy. NMES-often targeted paraspinal or multifidus muscles-has smaller evidence base; recent trials examine NMES as an adjunct to exercise or motor-control training with promising but preliminary results for improving muscle activation and short-term outcomes. Practical use: consider TENS for short-term symptom relief and NMES when the goal is to restore specific muscle activation (e.g., multifidus) alongside active exercise (Mervyn J Travers, 2020). 6.4 McKenzie Method (Mechanical Diagnosis and Therapy) The McKenzie method (MTD) is a patient-driven assessment and repeated-movement/positional exercise approach intended to centralize radicular symptoms and guide self-management. Systematic reviews indicate MDT can reduce pain for some patients with mechanical LBP and may perform similarly to other exercise-based approaches for certain outcomes. Predicting which patients will respond remains imperfect; MDT is most appropriate when directional preference/centralization is demonstrable on assessment supports it, and combine with exercise and education as needed (Fayez Ibrahim Namnaqani 1, 2019). 6.5 Core stabilization (motor-control) exercises Core stabilization or motor-control programs target deep trunk muscles (transversus abdominis, lumbar multifidus) with the aim of improving segmental control and reducing pain recurrence. Systematic reviews find evidence that stabilization exercises can reduce pain and disability in the short to medium term; however, several trials show that stabilization is not consistently superior to general exercise when followed long term. Clinically, motor-control or recurrent episodes-but should be embedded within a progressive, function-focused program (Zachary Smrcina 1, 2022). 6.6 Hydrotherapy, Pilates, Yoga (optional adjuncts) Aquatic (hydrotherapy) exercise leverages buoyancy loading to enable movement and graded strengthening – systematic reviews and trials report improvements in pain and function for chronic LBP, particularly when land-based exercise is limited by pain. Mind-body modalities such as Pilates and yoga also show modest benefits in pain and function in several RCTs and reviews; effects are generally comparable to other active exercise programs an may be particularly helpful for adherence, proprioception, and psychosocial aspects of chronic pain. These modalities are best viewed as adjuncts or alternative formats for exercise prescription, chosen according to patient preference, comorbidities, and access (Ji Ma 1, 2022). 7. Conclusion Physiotherapeutic management of low back pain is a fundamental component of conservative treatment and offers meaningful improvements in pan, mobility, and function. The literature consistently supports exercise therapy as the cornerstone of intervention, while manual therapy, core stabilization, and other targeted techniques can enhance clinical outcomes when applied appropriately. Although adjunctive modalities such as electrotherapy, hydrotherapy, Pilates and yoga demonstrate selective benefits. Their effectiveness depends on patient characteristics and integration with active treatment strategies. The heterogeneity of studies suggests that personalization of care, rather than a single standardized approach, yields the best results. 8. Recommendations Based on the reviewed evidence, several practice-oriented recommendations emerge: 1. Prioritize active treatment: Exercise therapy should form the basis of physiotherapeutic intervention for both acute and chronic LBP. XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 52 2. Use multimodal approaches: Combining exercise with manual therapy, stabilization training, and education enhances clinical outcomes. 3. Apply adjunct modalities selectively : TENS, NMES, hydrotherapy, Pilates, yoga should be used to support, not replace, active rehabilitation. 4. Individualize care: Patient-specific factors, including pain duration, functional limitations, and directional preference, should guide treatment selection. 5. Integrate education and self-management: Long-term improvement is more likely when patients are involved in their recovery and understand their condition. References 1. Airaksinen, O. B.-M. (2006). Chapter 4: European guidelines for the management of chronic nonspecific low back pain. European Spine Journal,. 2. Artus M, v. d. (2014). The clinical course of low back pain:. BMC Musculoskelet Disord. 3. Balagué, F. M. (2012). Non-specific low back pain. The Lancet, 482–491. 4. Collaborators., G. 2. (2018). Global, regional, and national incidence, prevalence, and years lived with disability for 354 diseases and injuries for 195 countries and territories, 1990-2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet Lond Engl., 1789–1858. 5. Da C Menezes Costa L, M. C. (2012). The prognosis of acute. CMAJ, 184:E613-24. 6. da Silva T, M. K. (2019). Recurrence of low back pain is common: a. J Physiother . 7. Delitto, A. G. (2012). Low back pain: Clinical practice guidelines linked to the International Classification of Functioning, Disability, and Health. Journal of Orthopaedic & Sports Physical Therapy. 8. Dieleman JL, C. J. (2020). US health care spending by payer and health condition. JAMA, ;323:863-84. 9. E., V. (1997). The epidemiology of low back pain in the rest of the world. A review of surveys in lowand middle-income countries. . Spine., 1747–1754. 10. Edwards J, H. J. (2017). Prevalence of low back pain in emergency settings: a systematic review and meta-analysis. . BMC Musculoskelet Disord., 18(1):143. 11. Excellence., N. I. (n.d.). Low back pain and sciatica in over 16s: Assessment and management (NG59). 12. Fayez Ibrahim Namnaqani 1, A. S. (2019). The effectiveness of McKenzie method compared to manual therapy for treating chronic low back pain: a systematic review. J Musculoskelet Neuronal Interact., 492-499. 13. Griffin D.W., H. D. (2012). Do patients with chronic low back pain have an altered level and/or pattern of physical activity compared to healthy individuals? A systematic review of the literature. Physiotherapy. , 13-23. 14. Hoy D., B. P. (2010). The Epidemiology of low back pain. Best Pract Res Clin Rheumatol., 769– 781. 15. Hoy, D. B. (2014). A systematic review of the global prevalence of low back pain. Arthritis & Rheumatology, 2028–2037. 16. Ji Ma 1, #. T. (2022). Effect of aquatic physical therapy on chronic low back pain: a systematic review and meta-analysis. BMC Musculoskelet Disord. . 17. Jill A Hayden 1, ,. J. (2021 Sep 28). Exercise therapy for chronic low back pain. Cochrane Database Syst Rev. 18. Knezevic N.N., C. K. (2021). Low back pain. Lancet Lond Engl., 78–92. 19. Koes, B. W. (2010). Diagnosis and treatment of low back pain. BMJ. 20. Kongsted A, K. P. (2016). What have we learned from ten years of trajectory research in low back pain? BMC Musculoskelet Disord. 21. Kreiner, D. S. (2014). An evidence-based clinical guideline for the diagnosis and treatment of lumbar disc herniation with radiculopathy. The Spine Journal, 180–191. 22. Kwon B.K., R. D. (2011). Systematic review: occupational physical activity and low back pain. Occup Med Oxf Engl., 541-548. 23. Lanier D.C., S. P. (1988). Clinical predictors of outcome of acute episodes of low back pain. . J Fam Pract., 483–489. 24. Maher, C. U. (2017). Non-specific low back pain. The Lancet. 25. Manchikanti L., S. V. (2014). Epidemiology of low back pain in adults. Neuromodulation J Int Neuromodulation Soc. . XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 53 26. Maniadakis N, G. A. (2000). The economic burden of back pain in the UK. Crossref PubMed., 84(1):95–103. 27. Mark Wilhelm 1, ,. J. (2023). The combined effects of manual therapy and exercise on pain and related disability for individuals with nonspecific neck pain: A systematic review with meta-analysis. J Man Manip Ther., 393-407. 28. Mervyn J Travers, N. E. (2020). Transcutaneous electrical nerve stimulation (TENS) for chronic pain: the opportunity to begin again. Download PDF. 29. Mervyn J TraversNeil E O'Connell, P. T. ( 22 April 2020). Transcutaneous electrical nerve stimulation (TENS) for chronic pain: the opportunity to begin again. Download PDF. 30. Roffey D.M., B. A. (2013). Obesity and low back pain: is there a weight of evidence to support a positive relationship? Curr Obes Rep. , 241–250. 31. S., B. (2015). Economic impact of musculoskeletal disorders (MSDs) on work in. Best Pract Res Clin Rheumatol, ;29:356-73. 32. Steffens D., F. M. (2015). What triggers an episode of acute low back pain? A case-crossover study. . Arthritis Care Res., 403–410. 33. Yalew E.S., A. K. (2022). Low back pain and its determinants among wait staff in Gondar town, North West Ethiopia: a cross-sectional study. . Front Pain Res Lausanne Switz. . 34. Zachary Smrcina 1, S. W. (2022). A Systematic Review of the Effectiveness of Core Stability Exercises in Patients with Non-Specific Low Back Pain. Int J Sports Phys Ther., 766-774. XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 54 Conclusions Autoimmune thrombocytopenias are disorders that can be fatal for the patient depending on their form. For this reason, rapid diagnosis and differentiation between them is important for the early intervention with appropriate treatment. Keywords: Thrombocytopenia, PTT, PTI, a-PTT, differential diagnosis, antibodies, hemorrhage, petechiae, ecchymosis, organ failure, thrombosis, corticosteroids, immunosuppressants References 1. Rodeghiero F, Stasi R, Gernsheimer T, Michel M, Provan D, Arnold DM, Bussel JB, Cines DB, Chong BH, Cooper N, Godeau B, Lechner K, Mazzucconi MG, McMillan R, Sanz MA, Imbach P, Blanchette V, Kühne T, Ruggeri M, George JN. Standardization of terminology, definitions and outcome criteria in immune thrombocytopenic purpura of adults and children: report from an international working group. Blood. 2009 Mar 12;113(11):2386-93 2. Zitek T, Weber L, Pinzon D, Warren N. Assessment and Management of Immune Thrombocytopenia (ITP) in the Emergency Department: Current Perspectives. Open Access Emerg Med. 2022;14:25-34. AUTOIMMUNE THROMBOCYTOPENIAS. DIFFERENTIAL DIAGNOSIS Dorina Ruci Prof. Asc. Vilson Ruci Prof. Asc. Krisli Serani Dr. Arsel Dizdari Dr. Vasilika Gjika MD Entela Shkodrani Prof. Asc.; Rheumatology Department, University Medical Center of Tirana "Mother Teresa" Abstract Autoimmune thrombocytopenias are classified as hematological disorders characterized by peripheral thrombocytopenia, mainly caused by the presence of antiplatelet autoantibodies. Patients can present with different clinical manifestations ranging from mild cutaneous and mucosal hemorrhages to massive life-threatening organ involvement. The main pathologies included in the differential diagnosis of autoimmune thrombocytopenias are: PTI (immune thrombocytopenic purpura), PTT (thrombotic thrombocytopenic purpura) and aPTT (acquired thrombotic thrombocytopenic purpura). PTI is caused by the presence of IgG antibodies against platelet membrane glycoproteins IIbIIIa. It is clinically characterized by the presence of petechiae, papules, ecchymoses mainly in the extremities, gingivorrhagia, menorrhagia in women. Severe forms with cerebral hemorrhage or gastrointestinal involvement are rare. PTT is caused by a congenital deficiency or decreased activity of the enzyme ADAMS13. It is characterized by the pentad: thrombocytopenia, fever, hemolytic anemia, renal involvement and neurological involvement. aPTT is caused by an acquired deficiency of the ADAMS13 enzyme caused by the presence of inhibitory autoantibodies. It is characterized by systemic microvascular thrombosis which leads to deep thrombocytopenia, hemolytic anemia and even organ failure. The differential diagnosis of these diseases presents a challenge due to their similar clinical manifestations. An important element in their differentiation are laboratory evaluations, mainly hematological tests, as well as the immunological ones to assess the presence of specific antibodies. Treatment in most cases is similar and includes corticosteroids as the first line of treatment, plasmapheresis, intravenous immunoglobulins and immunosuppressants. XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 55 3. Cines DB, Bussel JB, Liebman HA, Luning Prak ET. The ITP syndrome: pathogenic and clinical diversity. Blood. 2009 Jun 25;113(26):6511-21. 4. Coppo P, Cuker A, George JN. Thrombotic thrombocytopenic purpura: Toward targeted therapy and precision medicine. Res Pract Thromb Haemost. 2019 Jan;3(1):26-37. 5. Cox EC. Thrombotic thrombocytopenic purpura: report of three additional cases and a short review of the literature. J S C Med Assoc. 1966 Dec;62(12):465-70. 6. Swart L, Schapkaitz E, Mahlangu JN. Thrombotic thrombocytopenic purpura: A 5-year tertiary care centre experience. J Clin Apher. 2019 Feb;34(1):44-50. XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 56 Pedagogical sciences 1. Introduction The topic is highly important and relevant, as the education a child receives in the early stages of life plays a crucial role in their future development. The relevance of this topic is based on the fact that early childhood education serves as the foundation for an individual’s lifelong learning, personality formation, and social adaptation. Numerous studies in pedagogy and psychology confirm that the experiences and education a child gains between the ages of 0 and 6 have a lasting impact on their cognitive, emotional, and moral development. Therefore, ensuring access to quality preschool education is not only essential for the well-being and development of each child but also for the sustainable progress of society as a whole. In the context of Azerbaijan, the modernization of preschool education and state policies aimed at expanding its accessibility make this issue even more significant and timely. Concept of Preschool Education THE ROLE AND IMPORTANCE OF PRESCHOOL EDUCATION IN AZERBAIJAN Fakhriya Aslanova “MƏKTƏBİM” Teaching and Learning Center in Baku Branch, Azerbaijan AZERBAYCAN’DA OKUL ÖNCESİ EĞİTİMİN ROLÜ VE ÖNEMİ Fakhriya Aslanova “MƏKTƏBİM” Öğretim ve Öğrenim Merkezinin Bakü Şubesi (Azerbaycan) Abstract Preschool education in Azerbaijan plays a critical role in the holistic development of children, encompassing cognitive, social, emotional, and physical growth. This stage prepares children for formal schooling by enhancing their foundational skills in language, mathematics, and problemsolving. State policies and programs, including the 2013 order of the Ministry of Education, the 2014 Presidential Decree on the Development of Preschool Institutions, and the 2016 Cabinet decision on accreditation, have contributed to improving the quality and accessibility of preschool education. These initiatives aim to expand preschool coverage, modernize infrastructure, train qualified staff, and ensure equitable opportunities for all children. Research indicates that children who attend preschool demonstrate better academic performance, stronger social skills, and enhanced self-confidence. Therefore, preschool education is not only a preparatory stage for formal schooling but also a crucial factor in fostering lifelong learning, national identity, and sustainable societal development in Azerbaijan. Özet Azerbaycan’da okul öncesi eğitim, çocukların bilişsel, sosyal, duygusal ve fiziksel gelişimini kapsayan bütüncül bir rol oynar. Bu dönem, çocukların dil, matematik ve problem çözme gibi temel becerilerini geliştirerek onları ilköğretime hazırlar. 2013 tarihli Eğitim Bakanlığı emri, 2014 tarihli Okul Öncesi Kurumların Geliştirilmesine Dair Cumhurbaşkanlığı Kararnamesi ve 2016 tarihli Bakanlar Kurulu kararı gibi devlet politikaları ve programları, okul öncesi eğitimin kalitesini ve erişilebilirliğini artırmaya katkı sağlamıştır. Bu girişimler, okul öncesi kapsama alanını genişletmeyi, altyapıyı modernize etmeyi, nitelikli personel yetiştirmeyi ve tüm çocuklar için eşit fırsatlar sağlamayı hedeflemektedir. Araştırmalar, okul öncesi eğitime katılan çocukların akademik başarılarının daha yüksek, sosyal becerilerinin daha güçlü ve özgüvenlerinin daha yüksek olduğunu göstermektedir. Bu nedenle, okul öncesi eğitim sadece ilköğretime hazırlık aşaması değil, aynı zamanda ömür boyu öğrenmeyi, ulusal kimliğin gelişimini ve Azerbaycan’da sürdürülebilir toplumsal kalkınmayı destekleyen kritik bir faktördür. Keywords: Preschool education, Azerbaijan, teaching, learning, decisions, decrees Anahtar kelimeler: okul öncesi eğitim, Azerbaycan, öğretim, eğitim, kararlar XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 63 The positive impact of the digital route also manifests itself in emotional and psychological aspects. The ability to move at one's own pace reduces anxiety, especially when completing speaking and writing tasks, where students often experience a fear of making mistakes. The digital environment provides a safe learning space, where mistakes are perceived not as failures but as part of the self-improvement process. This promotes a positive attitude toward language and confidence in one's own communicative abilities. In terms of its social and communicative impact, the digital route transforms interactions between students and instructors. The shift from a frontal to an individualized model doesn't lead to isolation; on the contrary, the digital environment enhances collaboration. Collaborative assignments, group projects, and discussions in LMS forums create new models of academic communication, where students act not as passive listeners but as communication partners. The instructor, in turn, acts as a moderator and facilitator, helping students analyze their results and develop their own strategies. In this model, communication takes on a more equal and dialogical nature [5, p. 2]. The digital route also promotes the development of critical and analytical thinking. Automated feedback and access to personal progress data instill in students the habit of objective self-assessment, a necessary element of a mature academic personality. Students begin to perceive mistakes as part of the process, analyze the causes of difficulties, and seek ways to overcome them. This leads to the emergence of a new type of academic responsibility, based on an awareness of their personal role in achieving educational goals. For teachers, the digital route becomes not only a means of organizing the educational process but also a tool for diagnostics and pedagogical analytics. Using student activity data, teachers can adjust the course, identify weaknesses, adapt methods, and improve the quality of interaction. This system increases the transparency of pedagogical decisions and builds trust among participants in the educational process. Overall, the pedagogical effects of the digital route in English language teaching reflect a profound transformation in educational culture. The relationship between learner, teacher, and technology is no longer hierarchical but rather partnership-based. A new educational model is emerging—one that is human-centered, adaptive, and reflective, in which digital tools are not an end in themselves, but a means of developing autonomous individuals capable of learning, thinking, and communicating in a global digital world. • From Concept to Practice: Implementing a Digital Route in English Language Teaching The transition from a theoretical understanding of digital learning pathways to their practical implementation in English language teaching requires a comprehensive approach, including pedagogical design, technological support, and the willingness of educational stakeholders to collaborate. In recent years, digital learning pathways have become a central tool in university teaching methods, reflecting the trend toward personalized learning and increased student autonomy. The most illustrative examples are the implementation of this concept at Russian and international universities, where digital platforms have become not just a means of content delivery but a space for pedagogical interaction. In the teaching practice of Moscow State Pedagogical University (MSPU), the digital pathway concept is implemented through a combination of the Moodle and Google Classroom platforms. English language courses are structured as a sequence of modules united by a common logic of cognitive and communicative development. Each module contains goals, assignments, interactive exercises, and self-assessment tools. At the beginning of the course, students are offered a diagnostic test, based on which the system creates an individual pathway: more advanced students are given tasks of increased difficulty, while beginners are focused on basic topics and speech patterns. The instructor uses Learning Analytics reports to analyze progress, adjust materials, and provide personalized recommendations. This model creates a balance between freedom of choice and structured pedagogical support [1, p. 3]. At the University of Helsinki (Finland), a digital learning path is integrated into the "English for Academic Communication" course. It is implemented within the iSpring Learn platform, where each student builds a personalized trajectory for mastering academic English: from vocabulary and grammar to academic writing and public presentations. The system automatically adapts assignments based on the student's pace and level of success. The teacher does not set a uniform pace, but acts as a mentor, helping the student understand their results and adjust their learning strategy. A distinctive feature of the Finnish approach is its emphasis on reflection: after each module, the student completes a short report detailing which strategies were most effective and which require revision. This fosters a habit of self-analysis and strengthens critical thinking skills [2, p. 4]. XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 64 An example of the successful implementation of a digital learning path in Asian practice is Shanghai Normal University (China), which uses the integrated Tencent Classroom system. Here, the digital learning path for English is built on the principles of adaptability and collaboration. Courses are divided into micro-modules (microlearning), each of which represents a thematic block—grammar, listening, vocabulary, and intercultural communication. The AI system analyzes student behavior—task completion speed, error rate, and engagement level—and automatically creates an individualized learning path. Group work is organized through online discussions and virtual projects, where students create joint presentations and analyze English texts. The teacher acts as a facilitator, guiding the process and providing feedback through built-in analytical tools [3, p. 6]. An important area of practical implementation of the digital pathway is the use of visual planning tools such as Padlet, Trello, and Miro. These are used to design individual learning "roadmaps," where each stage reflects a specific topic, goal, and achievement criteria. For example, when studying the topic "Environmental Issues in Modern Society," students create their own pathways in Padlet: they post video resources, complete assignments, leave comments, and receive feedback from the instructor and classmates. This approach fosters creative thinking and enhances the communicative component of learning, transforming the course into a space for collaborative exploration and exchange of ideas [4, p. 2]. The effectiveness of implementing digital pathways is confirmed by empirical data from pedagogical research. For example, observations at universities using LMSs with personalization capabilities show that students experience a 23–28% increase in sustained learning motivation, while their level of independence in completing assignments increases by an average of 30%. Furthermore, improvements in academic performance and a significant increase in confidence in using English in oral and written communication have been recorded [5, p. 5]. Practical experience shows that the successful implementation of a digital route requires not only technical resources but also a change in teaching culture. Teachers become not just system operators but architects of the learning experience, designing routes, helping students understand the learning process, and transforming the digital space into a tool for self-development. It is the human factor that determines the effectiveness of technology: technology creates the environment, but it is the teacher and student who imbue it with meaning. The digital route to English language teaching, moving from concept to practice, is becoming not just an innovation, but a new format for interaction. It combines individualization, adaptability, and collaborative learning, creating the conditions for developing independent, responsible, and creative individuals capable of learning in a digital society and using language as a means of global communication. Conclusion The development of the concept of a digital learning path marks a shift from technology as a tool to technology as a form of pedagogical thinking. The experience of implementing personalized digital trajectories in English language teaching demonstrates that digitalization is not limited to the automation of the learning process; it opens the way to the development of a conscious, reflective, and humanistic approach to education. The digital path is becoming more than just an organizational model, but a space for intellectual and personal growth, where students don't just acquire knowledge, but build it based on their own experience, interest, and intrinsic motivation [2, p. 4]. The key achievement of the digital route is the creation of an educational environment where each student feels like a participant in the process, capable of setting goals, planning actions, and evaluating results. In this system, the teacher becomes the architect of the learning experience and a mentor, guiding the development of independence and critical thinking. This interaction gives rise to a new learning culture—a culture of trust, flexibility, and partnership between teacher and student [3, p. 7]. The digital learning path doesn't pit technology and personality against each other; on the contrary, it integrates them into a single system, where pedagogical logic and digital infrastructure work toward a common goal: developing fluent, reflective, and responsible language users. In the context of global educational transformation, such models are becoming the key to the sustainable development of language education, where technological innovation is driven not by pragmatics but by meaning. The prospect for further research lies in expanding the understanding of the digital pathway as part of a holistic educational ecosystem, where artificial intelligence, educational analytics, and humanistic methodology form a synthesis aimed at unlocking human potential. The digital pathway is not simply a path within the educational system, but a step toward a new type of education, in which knowledge becomes a process of self-discovery, and learning a form of personal maturity. Bibliography XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 65 1. Digital Learning Pathways: A Framework for Personalized Education. UNESCO Institute for Information Technologies in Education. — 2022. — URL: https://unesdoc.unesco.org/ark:/48223/pf0000382145 2. European Commission. Digital Education Action Plan 2021–2027. — Brussels: European Union, 2021. — URL: https://education.ec.europa.eu 3. Kangas, M., & Vesterinen, O. “Personalized Learning and Gamified Pathways in English Teaching.” // Education and Information Technologies. — 2021. — Vol. 26, No. 7. — P. 7891–7912. 4. Moodle HQ Documentation. Building Learning Pathways in LMS. — URL: https://docs.moodle.org 5. Siemens, G. Connectivism: A Learning Theory for the Digital Age. — International Journal of Instructional Technology & Distance Learning. — 2005. — Vol. 2, No. 1. XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 66 Technical sciences HYBRID HWOPSO-WOA FOR OPTIMAL FEATURE SELECTION IN DIABETIC RETINOPATHY PREDICTION Anamika Raj a, b Noor Maizura Mohamad Noora Rosmayati Mohemada Noor Azliza Che Mat a a Faculty of Computer Science & Mathematics, University Malaysia Terengganu, 21030 Kuala Nerus, Terengganu, Malaysia bDepartment of Computer Science, Applied College Al Mahala, 61421, King Khalid University, Saudi Arabia Abstract This research proposes a novel Hybrid Weighted Optimization-based Particle Swarm Optimization–Whale Optimization Algorithm (HWOPSO-WOA) for optimal feature selection in Diabetic Retinopathy (DR) classification. Conventional feature selection techniques often suffer from premature convergence and suboptimal search space exploration, resulting in redundant features and degraded model performance. The proposed HWOPSO-WOA combines the global exploration ability of Particle Swarm Optimization (PSO) with the adaptive exploitation strategy of Whale Optimization Algorithm (WOA), further enhanced with weighted opposition-based learning to maintain population diversity and accelerate convergence. Experimental evaluation on public DR datasets demonstrates a 25–30% reduction in feature dimensionality, with a 3–5% improvement in classification accuracy compared to standalone PSO, WOA, and other metaheuristic approaches. The hybrid algorithm exhibits improved stability and robustness across multiple runs, making it an efficient and scalable solution for high-dimensional medical data. Moreover, HWOPSO-WOA demonstrated faster convergence with lower computational cost compared to conventional hybrids, making it well-suited for real-time applications. Sensitivity analysis confirmed that the selected features captured key lesion characteristics, directly improving clinical interpretability of results. This contribution provides an optimized feature subset that enhances downstream classification models, thereby facilitating early DR detection and reducing diagnostic errors. Keywords: HWOPSO-WOA, Feature Selection, Metaheuristic Optimization, Diabetic Retinopathy, Swarm Intelligence, Hybrid Algorithms, Classification Accuracy XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 67 Introduction Pipeline infrastructure is a critical component in numerous industries, including oil, gas, and water supply. Ensuring its structural integrity is essential to prevent leaks, failures, and environmental hazards [1]. Traditional inspection methods, such as manual visual inspection, are time-consuming, laborintensive, and prone to human error [2]. In recent years, the integration of robotics and image processing techniques has provided an effective alternative for automated and precise pipeline inspection [3][4]. This chapter focuses on exploring various image processing techniques for detecting cracks in pipelines using an Arduino-based inspection robot. The study investigates five key approaches—Canny Edge Detection, Laplacian Operator, Morphological Operations, Otsu’s Thresholding, and Sobel Filter— implemented using Python 3 and the OpenCV library [5]. Each technique is analyzed in terms of its accuracy, computational speed, and efficiency in identifying crack features, including length, width, and skeleton structure [6]. We explored the conception and early phases of creating our pipe inspection robot in the previous chapter. We now delve further into its building, revealing elements intended to transform the effectiveness of pipe examination [3][4]. This chapter takes us on a thorough investigation of different image processing techniques. Every method is examined for how well it can identify the complex shapes of cracks. We begin a comparative analysis, comparing and contrasting each technique with its running process. We provide insight into the process of figuring out which strategy works best for our pipeline inspection projects through extensive testing. Among all these techniques, there are a few alternative methods under investigation. We must test them all to determine which best suits our needs because each has advantages and disadvantages. Ultimately, we want to find the method that best balances speed and accuracy to detect any possible problems in the pipes [6][7]. in the pipes. I will test five different techniques to see which one works best for detecting crack skeletons. ALGORITHM SELECTION AND OPTIMIZATION FOR IMAGE-BASED CRACK DETECTION E.N. İbrahimova Azerbaijan State University of Oil and Industry (ASOIU), Baku, Azerbaijan Abstract This chapter presents a comprehensive investigation into various image processing techniques for crack detection in pipeline inspection systems. The study utilizes Python 3 and the OpenCV library to implement and evaluate five key techniques: Canny Edge Detection, Laplacian Operator, Morphological Operations, Otsu’s Thresholding, and Sobel Filter. Each method was assessed in terms of accuracy, speed, and efficiency in detecting crack skeletons and structural parameters such as length and width. The results demonstrate that both Canny Edge Detection and Otsu’s Thresholding outperform other techniques in precision and computational performance. While Morphological and Laplacian methods showed limitations in noise reduction and accurate crack feature extraction, Otsu’s Thresholding proved particularly advantageous for its automated segmentation and high clarity in crack detection. Furthermore, real-time processing considerations revealed significant differences in execution times among techniques, with Otsu’s Thresholding emerging as the most efficient for practical implementation. These findings highlight the importance of optimizing image processing algorithms to achieve a balance between speed and accuracy in automated crack detection systems for pipe inspection. Keyword: Crack Detection, Image Processing, Canny Edge Detection, Otsu Thresholding, Pipeline Inspection XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 68 Figure 1. Opencv result of five image processing techniques from image 1 In our investigation, we employed the Python 3 OpenCV library to detect cracks. Each technique is characterized by distinct code commands. Common steps in image processing techniques include preprocessing by converting to grayscale and applying Gaussian blur for noise reduction, followed by edge detection using algorithms like Canny or Sobel. Thresholding methods such as global thresholding or Otsu's thresholding are then applied to segment the image into binary regions. For instance, Figure 1, 2, 3 and shows the implemenation of techniques. Figure 2. Opencv result of five image processing techniques from image 2 Original image Canny edge detection Laplacian Morphological operation Otsu’s thresholding Sobel filter Original image Canny edge detection Laplacian Morphological operation Otsu’s thresholding Sobel filter XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 69 Figure 3. Opencv result of five image processing techniques from image 3 Figure .4. Opencv result of five image processing techniques from image 4 From our analysis of crack images, it becomes evident that both Canny edge detection and Otsu thresholding methods exhibit superior accuracy in detecting crack parameters compared to Laplacian and morphological operation techniques. Specifically, we observed that morphological operations are limited in their ability to effectively reduce noise levels. Moreover, the Laplacian technique demonstrates inadequacies in accurately detecting crack features. As for image processing techniques, Otsu thresholding strikes me as particularly useful due to its ability to simplify the process of finding the optimal threshold for image segmentation. Canny edge detection stands out in its accuracy in identifying edges while minimizing false detections. In my experience applying these techniques to crack images, I've noticed variations in computational speed, complexity, and their ability to deliver the desired results in real-time applications. The important consideration is understanding the speed at which various techniques can detect cracks in images. To accomplish this, we utilize the `time.time()` function in OpenCV. By timing each technique's execution, we can identify which ones are faster at identifying cracks. This knowledge helps in choosing the most efficient technique for applications where speed is critical (Figure 7). Original image Canny edge detection Laplacian Morphological operation Otsu’s thresholding Sobel filter Original image Canny edge detection Laplacian Morphological operation Otsu’s thresholding Sobel filter XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 70 Figure 5. Implementing execution time of techniques In the provided benchmarks, I evaluated the computational speed of various image processing techniques using OpenCV (Figure 4.6). These techniques include Canny Edge Detection, Laplacian Operator, Morphological Operations, Otsu Thresholding, and Sobel Filter. The observed speeds indicate how quickly each technique processed the images. Some techniques, such as Canny Edge Detection and Otsu Thresholding, were notably faster than expected. Figure 6. Computing speed of 1) Sobel operator, 2) Morphological operation, 3) Laplacian, 4) Canny Edge, 5) Otsu’s threshold Figure .7. Speeds of each technique in Python terminal 1 2 3 4 5 0.0219 0.0109 0.004 0 0.0039 0.001 9 XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 71 Each technique exhibits distinct capabilities in detecting various parameters such as the length, width, and depth of cracks. Notably, both the Sobel operator and morphological methods demonstrate limitations in noise reduction, resulting in the presence of numerous features in the output. Similarly, the Laplacian technique fails to provide an accurate depiction of the crack skeleton. In contrast, Canny edge detection and Otsu's thresholding emerge as superior options, effectively detecting both the length and skeleton of cracks. Figure 8. Crack detecting accuracy proportion These findings imply that the chosen image processing techniques are suitable for the intended applications. Overall, the observed speeds align well with expectations, indicating effective performance for the selected image processing tasks. And I also concluded that every techniques show different results on their crack feature detections which I illustrated in Figure 4.8. According to practice, Otsu's thresholding technique is used more often, and I have determined in practice that I achieved these results. This technique detects cracks very clearly and faster than other used techniques, and I will take advantage of this in my crack detection process. References 1. Lohakare, P. H. (2025). Review of Crack Detection System for Industrial Pipe. International Journal of Structural and Civil Engineering Research, 14(3), 7818. Retrieved from https://www.ijsat.org/papers/2025/3/7818.pdf 2. Yuan, Q. (2024). A Review of Computer Vision-Based Crack Detection Methods. Remote Sensing, 16(16), 2910. https://doi.org/10.3390/rs16162910 3. Giulietti, N. (2025). Automated Vision-Based Concrete Crack Measurement System. Automation in Construction, 140, 103489. https://doi.org/10.1016/j.autcon.2022.103489 4. Shahin, M. (2024). Improving the Concrete Crack Detection Process via a Vision Transformer and Image Enhancement Detectors. Journal of Civil Structural Health Monitoring, 14(2), 123-135. https://doi.org/10.1007/s13349-024-00429-2 5. Muhi, O. S. (2025). Improving Oil Pipeline Surveillance with a Novel 3D Drone Simulation and CNN-Based Crack Detection. Journal of Pipeline Science and Engineering, 10(1), 1-14. https://doi.org/10.1007/s44196-025-00818-3 35% 25% 20% 15% 5% Otsu's thresholding Canny edge detection Sobel filter Morphological operation Laplacian operation XVII INTERNATIONAL SCIENTIFIC CONFERENCE. BEIJING. CHINA. 07-08.10.2025 72 Introduction Volume compaction of transmitted message without loss of its information content for a recipient was always topical issue at the processing of data in the radio engineering systems. Importance of this problem increases at the necessity to process heavy amounts of data in tight schedule. It acquires special relevance due to the sharp increase in the processed data volume as well as in need to more detailed study of measured processes. This requirement, of course, leads to the considerable complication in data transmission systems and radio-technical communication channels. Solution of the problems is possible at using of so-called data compaction methods. Normally, the broadband processes which constitute only (10-30) % of the general nomenclature for measured parameters but load transmitting channel on (60-80) % occupy a dominant position in the spectrum for measured parameters. Thus, the problem on compaction of broadband signals becomes particularly relevant. For compaction of broadband signal can use noninvertible data compaction methods which concluded in determination on the receiving side of information and measuring systems for probabilistic characteristic evaluation of the measured random processes and their transmission on communication channels. Methods Difficulties in then on invertible compaction methods are considerably higher, despite the fact that their effectiveness in terms of data compression is much higher. This is because data processed on the board very often present the sole implementation of non-stationary random processes in the absence of priori data on the form of distribution function. Modern mathematical statistics does not have methods to evaluate probabilistic characteristics of such processes. STATISTICAL INFORMATION PROCESSING IN RADIO ENGINEERING SYSTEMS Yesmagambetov Bulat-Batyr professor, Tashenev University, Shymkent, Kazakhstan Botayeva Saule associate professor, Tashenev University, Shymkent, Kazakhstan Kenzhebayeva Ulzhan senior lecturer, Tashenev University, Shymkent, Kazakhstan Adilet Makhanbetov master, M.Auezov South Kazakhstan University, Shymkent, Kazakhstan Sultanova Gulbanu master, M.Auezov South Kazakhstan University, Shymkent, Kazakhstan, Shymkent, Kazakhstan Tursumbayeva Adiya master, M.Auezov South Kazakhstan University, Shymkent, Kazakhstan Abstract Very often in radio engineering systems (for example rocket-and-space technics), the measured data represents a broadband non-stationary random process. The use of traditional processing methods that employ cyclic sampling results in high computational costs and large memory requirements, especially when dealing with onboard radio telemetry systems. As a rule, when processing random processes in such systems, restoration of the original implementation on the receiving end is not required, and processing involves calculating probability characteristics. In this case, data processing has a number of features. First, random processes are always represented by a single implementation. Second, a priori knowledge of the probabilistic properties of the measured random process is not always possible. Third, there is a need for real-time processing, which predetermines the use of fast processing methods. The article discusses the possibilities of applying methods of nonparametric decision theory to estimate the probabilistic properties of nonstationary broadband random processes in radio engineering systems. Keywords: radio engineering systems, random process, data compaction, nonparametric methods, order statistic, stationarity hypothesis.