EMERGING TRENDS IN ARTIFICIAL INTELLIGENCE (AI) -DRIVEN OPERATIONS: A BIBLIOMETRIC ANALYSIS

Copyright © Author(s) 2025. All Rights Reserved. Published by GLOBAL PUBLICATION HOUSE. | Int. J. Applied Management Science GPH-International Journal of Applied Management Science (e-ISSN 3050-9688 | Open Access | Peer-Reviewed) Article ID: gph/ijams/2025/2124 EMERGING TRENDS IN ARTIFICIAL INTELLIGENCE (AI) -DRIVEN OPERATIONS: A BIBLIOMETRIC ANALYSIS Dearielyn C. Maskarino, MEIE Affiliation: Chairperson, Industrial Engineering Department Palompon Institute of Technology, Palompon, Leyte, Philippines DOI: https://orcid.org/0009-0000-7106-5352 -------------------------------------------------------- Volume: 05 | Issue: 09 | September 2025 | Pages: 14–29 DOI: 10.5281/zenodo.17356463 | www.gphjournal.org Publisher: Global Publication House -------------------------------------------------------- How to cite: Maskariño, D. (2025). EMERGING TRENDS IN ARTIFICIAL INTELLIGENCE (AI) - DRIVEN OPERATIONS:. GPH-International Journal of Applied Management Science, 5(9), 14-29. https://doi.org/10.5281/zenodo.17356463 Page 14 of 29 3050-9688 EMERGING TRENDS IN ARTIFICIAL INTELLIGENCE (AI) -DRIVEN OPERATIONS Volume 05 Issue No 09 (2025) Open Access: https://gphjournal.org/index.php/ams EMERGING TRENDS IN ARTIFICIAL INTELLIGENCE (AI) -DRIVEN OPERATIONS: A BIBLIOMETRIC ANALYSIS Abstract Background Artificial Intelligence (AI) has become a transformative force in operational domains, reshaping processes in manufacturing, logistics, scheduling, and cloud-based systems. The rapid proliferation of research output, particularly within the 2025–2026 period, underscores the need for a systematic bibliometric assessment to elucidate emerging thematic trajectories, intellectual structures, and influential contributors in AI-driven operations. Methods A quantitative bibliometric design was employed using the Biblioshiny platform, underpinned by the Bibliometrix R package. Bibliographic data were sourced from Scopus and restricted to publications dated 2025–2026 that explicitly addressed AI applications in operational contexts. The analysis integrated performance indicators such as publication productivity and citation impact with science mapping techniques, including co-authorship analysis, keyword co-occurrence, thematic clustering, and network centrality metrics. Results Findings reveal a pronounced temporal concentration of publications in 2025, indicative of a hyperaccelerated research front. China emerged as the predominant contributor, with South China University of Technology and other leading institutions demonstrating the highest output. Thematic mapping identified three major clusters reinforcement learning, scheduling algorithms, and smart manufacturing and a smaller emergent cluster on fabrication. Strong inter-thematic linkages highlight the convergence of AI methodologies with operational optimization and Industry 4.0 applications. Owing to the recency of the dataset, traditional citation counts were minimal; thus, PageRank and network-based metrics provided more meaningful indicators of early influence. Several recent publications demonstrated notable structural impact within the emerging knowledge network. Conclusion AI-driven operations research is characterized by rapid expansion, thematic convergence, and significant regional concentration, particularly within Chinese institutions. Reinforcement learning, scheduling algorithms, and smart manufacturing constitute the intellectual core of the field, reinforced by advances in cloud and edge computing. In the context of an emergent research landscape, networkbased impact measures are more appropriate than conventional citation metrics. The findings indicate a swift transition from theoretical exploration to applied innovation, necessitating continued monitoring, interdisciplinary collaboration, and strategic policy and industry engagement. Keywords Artificial intelligence, operations research, reinforcement learning, scheduling algorithms, smart manufacturing, bibliometric analysis. 15 Maskariño, D. (2025). EMERGING TRENDS IN ARTIFICIAL INTELLIGENCE (AI) -DRIVEN OPERATIONS:. GPH-International Journal of Applied Management Science, 5(9), 14-29. https://doi.org/10.5281/zenodo.17356463 © GPH-International Journal of Applied Management Science | www.gphjournal.org Introduction The integration of Artificial Intelligence (AI), particularly machine learning and deep learning, is reshaping traditional operational systems across multiple sectors (Soori, Arezoo, & Dastres, 2023). Industries such as manufacturing, logistics, supply chain management, and distributed computing increasingly rely on AI to enhance efficiency, responsiveness, and decision-making autonomy. These technologies enable intelligent scheduling, dynamic resource allocation, predictive maintenance, and automation in complex environments, including cloud infrastructures and cyberphysical systems. By embedding adaptive capabilities into operational processes, AI drives not only efficiency gains but also resilience, flexibility, and scalability in rapidly evolving industrial settings (Sundaramurthy, Ravichandran, Inaganti, & Muppalaneni, 2022). As AI becomes more deeply embedded in operational contexts, bibliometric analysis has gained prominence as a methodological tool for mapping the structure, growth, and intellectual evolution of the field. Previous scientometric studies, such as the analysis of smart cities and sustainable development (2015–2018), have demonstrated the value of mapping intellectual structures to understand technology-driven research frontiers (Sanico, M.F., 2025). Bibliometric techniques provide a systematic framework for identifying publication trends, influential authors and institutions, collaboration networks, and emerging thematic patterns (Zucolotto, Yamane, & Siman, 2022). These analyses offer strategic insights for researchers, policymakers, and industry stakeholders seeking to monitor scientific developments, allocate resources, and identify research gaps (Skute, Zalewska-Kurek, Hatak, & de Weerd-Nederhof, 2017). In the context of rapidly advancing AI applications, bibliometric mapping enables the identification of emerging subfields with high transformative potential. Current developments in AI-driven operations are marked by the convergence of intelligent algorithms with real-time data processing, enabling systems to operate autonomously in dynamic environments (Ekundayo, 2024). Reinforcement learning (RL) has emerged as a critical methodological approach for adaptive decision-making in domains such as logistics, scheduling, and inventory management. Its capacity for continuous learning through environmental feedback makes it well-suited to high-variability contexts. Simultaneously, the integration of AI with the Internet of Things (IoT) and cyber-physical systems has advanced the development of smart manufacturing ecosystems, where automation, data exchange, and self-optimization are central features (Dave, 2023). These technological shifts signal a transition toward predictive and prescriptive analytics as core components of next-generation operational systems. The integration of AI into cloud-based and edge-computing infrastructures has further expanded its applicability across sectors such as healthcare, transportation, and energy (Kumar, 2022). Interdisciplinary collaboration is accelerating the development of hybrid approaches that combine classical optimization with AI-driven decision models. The increasing emphasis on explainable AI (XAI) reflects a parallel need for transparency, accountability, and ethical alignment in automated operational systems (N et al., 2024). These developments point to a rapidly evolving research landscape characterized by both technological innovation and cross-sector applicability. 16 EMERGING TRENDS IN ARTIFICIAL INTELLIGENCE (AI) -DRIVEN OPERATIONS Volume 05 Issue No 09 (2025) Open Access: https://gphjournal.org/index.php/ams Given this dynamic context, the present bibliometric study aims to provide a comprehensive and timely overview of the research landscape on AI-driven operations. Specifically, it seeks to: (1) analyze publication output and growth trajectories; (2) identify the most prolific and influential authors and institutions; (3) map dominant research themes and examine their interrelationships; and (4) highlight emergent and high-impact publications using both traditional and network-based citation metrics. By focusing on publications from 2025 to 2026, this study offers an early assessment of an accelerating research frontier and contributes to the strategic understanding of future research directions in AI-driven operational systems. Methods Study Design This study employed a quantitative bibliometric research design to systematically examine the intellectual structure and thematic evolution of AI-driven operations research. Bibliometric analysis was selected for its capacity to provide objective insights into publication trends, influential entities, collaboration patterns, and emerging conceptual domains. The approach combined performance analysis—focusing on productivity and impact metrics—with science mapping techniques that reveal structural and thematic relationships within the literature. The design is exploratory and descriptive, aligned with the principles of science mapping, and aimed at capturing both the breadth and depth of scholarly activity in this rapidly developing field. Data Source The primary data source for this study was the Biblioshiny platform, an R-based web interface built upon the Bibliometrix package. Biblioshiny facilitates advanced bibliometric analysis and visualization, enabling the examination of metadata derived from established bibliographic databases. Scopus was selected as the source database due to its comprehensive coverage of peerreviewed publications in engineering, computer science, technology, and related domains pertinent to AI and operations research. Extracted metadata included publication titles, authors, affiliations, keywords, source journals, citation counts, and publication years, providing a robust foundation for subsequent analyses. Search Strategy and Data Extraction A structured search strategy was implemented to identify relevant literature at the intersection of Artificial Intelligence and Operations Research. Boolean operators and controlled vocabulary were used to formulate the query, incorporating key terms such as “Artificial Intelligence,” “Machine Learning,” “Reinforcement Learning,” “Smart Manufacturing,” “Scheduling Algorithms,” and “Operational Optimization.” The search was limited to article titles, abstracts, and keywords to ensure thematic relevance and was restricted to publications from 2025 to 2026 to capture the emerging research front. 17 Maskariño, D. (2025). EMERGING TRENDS IN ARTIFICIAL INTELLIGENCE (AI) -DRIVEN OPERATIONS:. GPH-International Journal of Applied Management Science, 5(9), 14-29. https://doi.org/10.5281/zenodo.17356463 © GPH-International Journal of Applied Management Science | www.gphjournal.org The search was conducted in Scopus. Retrieved documents were exported in BibTeX format and uploaded into Biblioshiny for processing. The exported metadata included essential fields for performance indicators and network-based analyses, such as author names, institutional affiliations, keyword occurrences, citation data, and source titles. Both journal articles and conference papers were retained to reflect the multidisciplinary and fast-evolving nature of the field. Data Screening and Pre-processing Following extraction, the dataset underwent screening to ensure relevance and data quality. Duplicate entries, incomplete records, and publications outside the scope of AI-driven operational applications were removed. Only documents explicitly addressing AI techniques applied to operational processes were retained. Editorials, errata, and non-English publications were excluded to maintain consistency and methodological rigor. Preprocessing procedures were conducted to standardize metadata and enhance the reliability of network analyses. Author names, institutional affiliations, and keywords were normalized to address inconsistencies in spelling and formatting. Keyword harmonization was achieved through stemming and consolidation of variants (e.g., “optimising” and “optimization”). These steps were essential to minimize fragmentation and ensure accurate representation of co-authorship and keyword co-occurrence networks. The finalized dataset was then prepared for analysis within Biblioshiny. Data Analysis Data analysis comprised two major components: performance analysis and science mapping. Performance analysis examined publication productivity by authors, institutions, countries, and source journals, alongside citation-based metrics such as total citations, average citations per document, and h-index. Annual publication trends were also assessed to identify growth trajectories. Science mapping techniques were employed to explore the conceptual and collaborative structure of the field. Co-authorship analysis was conducted to map collaboration networks among authors, institutions, and countries. Keyword co-occurrence analysis was used to identify dominant themes and their interconnections, while thematic evolution mapping traced shifts in research priorities over time. Visualization tools within Biblioshiny, including thematic maps, collaboration graphs, and trend topic plots, were used to support interpretation. Together, these methods yielded a comprehensive understanding of the field’s structural dynamics and emergent research directions. Results The bibliometric dataset reveals a distinctly concentrated temporal pattern, with the majority of publications dated 2025 and extending minimally into 2026. This sharply contrasts with the cited references, which span from 1776 to 2020. This temporal asymmetry indicates that the present analysis is not retrospective but rather captures an emergent and rapidly accelerating research 18 EMERGING TRENDS IN ARTIFICIAL INTELLIGENCE (AI) -DRIVEN OPERATIONS Volume 05 Issue No 09 (2025) Open Access: https://gphjournal.org/index.php/ams front. The compressed timeframe underscores the immediacy of scholarly activity in AI-driven operations and reflects the rapid pace of innovation in the field. Author Productivity The analysis of author contributions demonstrates a high degree of concentration among leading scholars. WANG X is the most prolific contributor with 11 articles and a fractional authorship score of 2.2714. This is followed by LI X and LI Y, each with nine publications, with fractionalized contributions of 1.8551 and 2.7095, respectively. The use of fractional authorship provides a more accurate assessment of individual contributions within multi-author publications, which are prevalent in this collaborative domain. Table 1. Prolific Authors Authors Articles Articles Fractionalized WANG X 11 2.2714 LI X 9 1.8551 LI Y 9 2.7095 WANG L 9 1.6782 WANG Y 9 1.9119 LI H 8 1.8023 WANG J 8 1.3940 ZHANG X 8 1.3123 CHEN X 7 1.1388 LIU Y 7 1.2261 LIU Z 7 1.3190 ZHANG J 7 1.3916 ZHANG Y 7 2.1261 CHEN Y 6 1.1670 LI M 6 1.3000 ZHANG H 6 1.2713 19 Maskariño, D. (2025). EMERGING TRENDS IN ARTIFICIAL INTELLIGENCE (AI) -DRIVEN OPERATIONS:. GPH-International Journal of Applied Management Science, 5(9), 14-29. https://doi.org/10.5281/zenodo.17356463 © GPH-International Journal of Applied Management Science | www.gphjournal.org Institutional Contributions Institutional productivity is similarly concentrated, with Chinese universities and research institutions dominating the landscape. South China University of Technology ranks highest with 36 publications, followed by Islamic Azad University (32 articles) and Beihang University (22 articles). Additional leading contributors include Guizhou University, Guangdong University of Technology, and Wuhan University of Technology. The prominence of Chinese institutions suggests substantial national investment and research prioritization in AI-driven operational technologies. This trend highlights China's strategic positioning in advancing innovation and shaping global trajectories within this domain. Table 2. Influential Institutions Affiliation Articles SOUTH CHINA UNIVERSITY OF TECHNOLOGY 36 ISLAMIC AZAD UNIVERSITY 32 BEIHANG UNIVERSITY 22 GUIZHOU UNIVERSITY 19 GUANGDONG UNIVERSITY OF TECHNOLOGY 17 WUHAN UNIVERSITY OF TECHNOLOGY 17 AMERICAN COLLEGE OF CLINICAL PHARMACY 16 BEIJING UNIVERSITY OF TECHNOLOGY 16 CENTRAL SOUTH UNIVERSITY 16 UNIVERSITI PUTRA MALAYSIA 16 THE HONG KONG POLYTECHNIC UNIVERSITY 15 XI'AN JIAOTONG UNIVERSITY 15 CHONGQING UNIVERSITY 14 HUAZHONG UNIVERSITY OF SCIENCE AND TECHNOLOGY 14 TIANJIN UNIVERSITY 14 20 EMERGING TRENDS IN ARTIFICIAL INTELLIGENCE (AI) -DRIVEN OPERATIONS Volume 05 Issue No 09 (2025) Open Access: https://gphjournal.org/index.php/ams Thematic Clusters The thematic analysis revealed four distinct research clusters, three of which form the intellectual core of the field: reinforcement learning, smart manufacturing, and scheduling algorithms, with a fourth, smaller cluster focused on fabrication. Reinforcement learning emerged as the most influential thematic hub, with 1,343 associated keywords and the highest centrality scores (Callon Centrality 12.5388, Rank Centrality 4). Frequently linked with concepts such as deep reinforcement learning, machine learning, optimisation, decision making, and artificial intelligence, this cluster reflects its foundational role in methodological innovation across operational contexts. Smart manufacturing, comprising 501 keywords, is closely associated with terms like Industry 4.0, digital twin, flexible manufacturing systems, and predictive maintenance. Its centrality measures (Callon Centrality 5.4984, Rank Centrality 2) indicate its importance as a leading application domain for AI within industrial environments. Scheduling algorithms represent another major thematic pillar, with a frequency of 1,453 keywords and similarly high centrality (Callon Centrality 12.2394, Rank Centrality 3). Dominant terms such as resource allocation, cloud computing, task scheduling, and edge computing underscore its relevance in optimizing performance within distributed and cloudbased systems. A fourth and comparatively smaller cluster, fabrication, includes only 22 keywords and features niche topics such as forming and papermaking. Its low centrality and density suggest that it represents an emerging or specialized area within the broader scope of AI-driven operations research. Table 3. Research Themes with Centrality Metrics Cluster Keyword Frequency Centrality (Callon / Rank) Key Associated Terms Thematic Role Reinforcement Learning 1,343 12.5388 / 4 Deep reinforcement learning, machine learning, optimisations, decision making, artificial intelligence Foundational methodological hub Smart Manufacturing 501 5.4984 / 2 Industry 4.0, digital twin, flexible manufacturing systems, predictive maintenance Major application domain Scheduling Algorithms 1,453 12.2394 / 3 Resource allocation, cloud computing, task scheduling, edge computing Core thematic area addressing optimization 21 Maskariño, D. (2025). EMERGING TRENDS IN ARTIFICIAL INTELLIGENCE (AI) -DRIVEN OPERATIONS:. GPH-International Journal of Applied Management Science, 5(9), 14-29. https://doi.org/10.5281/zenodo.17356463 © GPH-International Journal of Applied Management Science | www.gphjournal.org Cluster Keyword Frequency Centrality (Callon / Rank) Key Associated Terms Thematic Role Fabrication 22 Low density and centrality Forming, papermaking Emerging/niche research area The interconnections across the three dominant clusters reveal not mere coexistence but a synergistic research ecosystem. Reinforcement learning functions as a core methodology that enables more advanced scheduling algorithms, which in turn support the optimization requirements of smart manufacturing systems. This alignment illustrates a systematic pipeline from algorithmic innovation to applies industrial deployment. Word Cloud Analysis A word cloud visualization was generated to provide a complementary overview of high-frequency terms within the dataset. The most prominent keywords such as reinforcement learning, scheduling algorithms, smart manufacturing, cloud computing, and optimization correspond closely with the dominant thematic clusters identified through keyword co-occurrence analysis. The visibility of additional terms such as digital twin, resource allocation, industry 4.0, and edge computing further illustrates the strong alignment between AI methodologies and their industrial and computational applications. While less structurally detailed than network-based approaches, the word cloud offers an accessible depiction of topic salience and reinforces the centrality of AI-driven operational innovation within the analyzed literature. 22 EMERGING TRENDS IN ARTIFICIAL INTELLIGENCE (AI) -DRIVEN OPERATIONS Volume 05 Issue No 09 (2025) Open Access: https://gphjournal.org/index.php/ams van Eck, N. J., & Waltman, L. (2014). Visualizing bibliometric networks. In Y. Ding, R. Rousseau, & D. Wolfram (Eds.), Measuring scholarly impact (pp. 285–320). Springer. https://doi.org/10.1007/978-3-319-10377-8_13 Venkatesh, V., Thong, J. Y. L., & Xu, X. (2020). Unified theory of acceptance and use of technology: A review and extension. MIS Quarterly, 45(1), 1–30. https://doi.org/10.25300/MISQ/2020/13842 Wang, J., Li, X., & Zhang, H. (2022). Resource allocation in edge–cloud environments: A reinforcement learning approach. IEEE Transactions on Cloud Computing, 10(2), 846–858. https://doi.org/10.1109/TCC.2019.2895718 Waltman, L. (2016). A review of the literature on citation impact indicators. Journal of Informetrics, 10(2), 365–391. https://doi.org/10.1016/j.joi.2016.02.007 Xu, L. D., Xu, E. L., & Li, L. (2018). Industry 4.0: State of the art and future trends. International Journal of Production Research, 56(8), 2941–2962. https://doi.org/10.1080/00207543.2018.1444806 Yan, E., & Ding, Y. (2009). Applying centrality measures to impact analysis: A co-authorship network analysis. Journal of the American Society for Information Science and Technology, 60(10), 2107–2118. https://doi.org/10.1002/asi.21128 Zhang, Y., Ren, S., Liu, Y., & Si, S. (2019). A big data analytics architecture for cleaner manufacturing and maintenance processes of complex products. Journal of Cleaner Production, 228, 976– 988. https://doi.org/10.1016/j.jclepro.2019.04.225 Zucolotto, P. V., Yamane, L. H., & Siman, R. R. (2022). Smart bibliometrics: An integrated method of science mapping and bibliometric analysis. Scientometrics, 127(1), 321–340. https://doi.org/10.1007/s11192-021-04173-7 29