Authors - Mainford Mutandavari, D. Hemavathi Abstract - Maize (Zea mays L.) is an essential staple produce for smallholder farmers in developing nations, yet Northern Corn Leaf Blight (NLB), Grey Leaf Spot (GLS), and Common Rust foliar diseases cause yield losses of 30–70%. Infection detection is done at advanced stages due to labor intensity resulting from the conventional disease monitoring methods. A Low-Cost Drone-Mounted Multispectral Imaging (LCDMI) framework for resource-constrained smallholder systems is presented in this paper, pairing a consumer-grade UAV with a five-band multispectral sensor. The vegetation-index features are fused with multispectral band data using a Spectral-Spatial Attention Vision Transformer (SSAViT) classifier and a Spectral-Constrained Synthetic Data Generation (SC-SDG) module addresses training-data scarcity. A hardware cost of USD1,940 is projected for field evaluations across twelve plots in Zimbabwe over two growing seasons yielding 95.8% detection accuracy, identifying diseases 7–12 days before visible symptom onset. A multi-label extension enables simultaneous classification of co-occurring infections. Georeferenced disease maps are delivered within 6.3 min/ha. With perhectare costs as low as USD2.10 on a scale, the economic analysis projects ROI within two seasons for cooperatives managing 50+ hectares.
Authors - IGN Oka Ariwangsa, Komang Widhya Sedana Putra P, Wayan Sri Maitri Abstract - The rapid adoption of artificial intelligence (AI) in higher education has transformed how students access information and engage in academic activities. While AI-powered technologies enhance efficiency and provide personalised support, their uncritical use may weaken independent reasoning and reduce meaningful social-academic participation. This raises concerns in digitally mediated environments where individuals must interpret complex information, evaluate uncertainty, and make informed judgments. Despite growing attention, most studies emphasise functional outcomes such as academic performance, overlooking the mechanisms through which AI-integrated teaching can foster deeper, more sustainable learning. This study examines how AI-aware pedagogy—defined as the intentional and reflective integration of AI in instructional design—enhances critical thinking through social-academic engagement. A quantitative approach was employed, involving 200 undergraduate students in Indonesia. Data were collected via structured questionnaires and analysed using Partial Least Squares Structural Equation Modeling (PLS-SEM). The findings reveal that AI-aware pedagogy has no significant direct effect on critical thinking. However, it significantly influences critical thinking indirectly through social-academic engagement. This indicates that higher-order thinking develops not merely through technological integration, but through socially embedded learning processes that encourage interaction, reflection, and evaluation. Theoretically, this study links digital pedagogy with cognitive and social learning processes. Practically, it highlights the need for AIsupported environments that foster critical evaluation and responsible decisionmaking under conditions of uncertainty. Future research should explore its applicability across contexts and its long-term cognitive implications.
Authors - Mary Diana C. Yamzon, Janelli M. Mendez Abstract - This study provides a data-driven analysis of the academic performance of Bachelor of Science in Office Administration (BSOAD) students at Tagbilaran City College from Academic Year 2021 to 2024, employing data mining clustering techniques to ascertain the five most challenging subjects. The study specifically aimed to: (1) construct and preprocess a dataset of pertinent academic attributes; (2) employ K-Means, K-Medoids, and Agglomerative Hierarchical Clustering algorithms to discern groupings of subject difficulty; (3) validate clustering results utilizing the Davies-Bouldin Index (DBI); and (4) develop evidence-based recommendations for curriculum enhancement and academic assistance. The analysis involved a dataset of 26,965 valid student grade records across 68 subjects, all of which were processed using RapidMiner Studio. The research utilized the Cross-Industry Standard Process for Data Mining (CRISP-DM) framework within the context of Educational Data Mining (EDM). The DBI for K-Means (DBI = 0.461; Excellent) and K-Medoids (DBI = 0.9145) were used to check the clusters, and the visual dendrogram was used to check the Agglomerative Hierarchical Clustering. All three algorithms consistently recognized OA113 Advanced Shorthand and OA111 Foundations of Shorthand as the two most challenging subjects in the program. The results offer statistically substantiated, evidence-based insights to facilitate curriculum evaluation, instructional enhancement, and the formulation of specialized academic intervention programs for BSOAD students.
Authors - Hussein P. El Sayed Ahmed, Ardee Joy T. Ocampo Abstract - Youth participation in local governance remains a persistent challenge despite institutional mechanisms designed to promote engagement. In the Philippines, the Sangguniang Kabataan (SK) serves as a formal platform for youth involvement in local decisionmaking; however, many SK programs continue to experience low participation, limited feedback integration, and repetitive activity design. This study presents EmpowerSK, a data-driven framework that leverages data mining techniques to enhance youth engagement in SK programs. Using structured survey data from 1,055 youth respondents aged 18–25 across the nine barangays of Alilem, Ilocos Sur, the study applies the Knowledge Discovery in Databases (KDD) framework, K-Means clustering, and sentiment analysis to transform raw feedback into governance intelligence. K-Means clustering (k=3) identified three statistically validated engagement profiles: Highly Active (61.6%), Moderately Involved (17.0%), and Disengaged (21.3%). Sentiment analysis of open-ended responses revealed appreciation (77.8% positive), diagnosis (73.2% negative), and aspiration (85.5% neutral-aspirational) as a coherent three-phase youth governance narrative. An overall weighted mean of 3.75 ("Very Good") across eleven Likert-scale items confirmed a critical institutional gap: Digital Engagement (4.14) significantly outpaced SK Support Initiatives (3.52), with SK Training recording the lowest item score (3.44). A five-pillar data-driven action plan—Awareness and Inclusion, Program Diversification, Digital Transformation, Capacity Building, and Monitoring and Evaluation—was developed, validated by SK officials, and aligned with SDG 4, 11, and 16. The findings demonstrate that freely available data mining tools can transform rural youth governance into an annually replicable, evidence-based participatory system.
Authors - Reynaldo F. Agunod, Janelli M. Mendez Abstract - Higher education institutions collect large volumes of student data but these are underutilized for institutional planning. This study applies the CRISP-DM framework to enrolment records of a freshman cohort of 1,916 students across four academic years (2021-2025) across 28 academic programs from a private higher education institution in Central Visayas, Philippines, to forecast institutional progression metrics using predictive analytics. Descriptive analytics and three predictive models were applied based on their suitability for the dataset with 3-4 data points, namely: Linear Regression, Holt-Winters Exponential Smoothing, and ARIMA. Six institutional performance metrics were analyzed: enrolment, retention, persistence, attrition, program shifts, and graduation. Key findings reveal a continuous 29.6% enrolment decline within the cohort, an im-proving retention and persistence profile, a program-shift surge largely due to migrations from Accountancy to Finance, and a rapidly increasing graduation rate. Linear Regression (OLS) was identified as the most effective forecasting model for the study’s single-cohort dataset.
Authors - Edimar J. Rato, Janelli M. Mendez Abstract - Student dropout has remained a major problem in all higher education institutions globally, including in the Philippines, where the total college dropout rate in the country was recorded at about 35.15% in the Academic Year 2023–2024. This study aimed to develop a predictive analytics model that identifies dropout and retention patterns among students of Tagbilaran City College to support evidence-based intervention strategies. offered by the school from Academic Year 2021-2024. The algorithms implemented for the supervised learning process include Random Forest and Gradient Boosting, while the algorithm for the unsupervised learning process is K-Means Clustering implemented using the RapidMiner Studio tool. Results revealed that both supervised models had a poor performance due to class imbalance issues as well as a small feature set; the Random Forest model had an accuracy of 59.59%, while it had an AUC of 0.575. The Gradient Boosting model had an accuracy of 60.51%, while it had an AUC of 0.508. The K-Means Clustering model had a good performance since it resulted in three interpretable student risk clusters: a moderate-risk group with a dropout rate of 27.3%, a highest-risk group with a dropout rate of 44.7%, and a lower-risk but larger group with a dropout rate of 41.9%. The Davies-Bouldin Index of 0.967 confirmed adequate cluster separation. The K-Means model demonstrated the most practical utility as an early-warning risk stratification tool applicable at the start of each academic year, forming the foundation of an evidence-based intervention plan to improve student retention at Tagbilaran City College.
Authors - Ni Made Prasiwi Bestari, Jonathan Jacob Paul Latupeirissa, Suryanto Nugroho, Iwan Adinugroho, Melati Budi Srikandi, Ayu Made Bianca Juarez Abstract - The Fourth Industrial Revolution has been transforming the global tourism industry, shifting toward a dynamic Tourism 4.0 ecosystem. Given that the adoption of AI is expected to increase the revenue of the tourism industry, it is necessary to conduct a Systematic Literature Review to fill the gap in empirical research on the relationship between technological innovation and long-term sustainability. Most studies on smart tourism from different perspectives, including tourist behavior, tourist service quality, innovation, and sustainability, focus on the "hardware construction" at the macro level and its implementation based on related policies, ignoring the psychological mechanisms affecting tourists' experiences at the micro level. This study aims to identify the key technological drivers, including AI, IoT, and computer vision, and their influence on operational innovation and Sustainable Development Goals. A total of 23 core manuscripts from 2020 to 2025 gathered from Scopus database were synthesized and analyzed based on PRISMA guidelines. The results showed that smart tourism technologies can greatly improve efficiency and enhance hyper-personalization. However, most current applications of smart tourism technologies do not take adequate account of social and environmental metrics. Also, many digital tourism strategies prioritize revenue over social inclusion. For the future of smart tourism destinations, frameworks such as Society 5.0 that integrate high-tech with the human touch of hospitality and tourism are needed. Destinations should also seek governance models that ensure long-term resilience by moving the focus away from infrastructure and toward "Smart People" initiatives and the development of standardized real-time sustainability metrics.
Authors - Ovia Rizvi, Sadman Kabir, Md. Tafshir Jaman Takib, Abir Sen Gupta, Sayra Islam Saki, S.M. Sayem Abstract - The global shift toward cashless payment systems has transformed financial transactions, yet adoption in developing countries such as Bangladesh remains limited. This study investigates the determinants of cashless payment adoption in Bangladesh by examining user perceptions and behavioral drivers. Drawing on survey-based evidence from 369 respondents, the PLS-SEM analysis identifies facilitating conditions, perceived security, initial trust and intrinsic motivation as the most influential factors shaping adoption. In contrast, digital financial literacy, social influence and IT innovation acceptance were found to have little impact, suggesting that peer effects and novelty alone do not encourage sustained use. Moreover, initial trust and intrinsic motivation showed significant mediating impact between the drivers and the adoption of cashless payment systems. The findings highlight the importance of robust infrastructure, strong security protocols and user-centric design in promoting digital financial inclusion. Policy implications emphasize collaborative efforts by regulators and service providers to expand infrastructure, enforce cybersecurity standards and foster user trust. These measures are critical for accelerating Bangladesh’s transition toward a secure and inclusive cashless society.
Authors - Aditya Nova Putra, Tri Wiyana, Setiani Putri Hendratno, Nora Fitriawati, Ida Bagus Putu Aditya Abstract - The world of social media marketing is shifting from traditional con-tent delivery to personalized solutions, algorithmic recommendation systems, AI generated content and Automated Customer Service Chat. While these technologies can increase relevance and responsiveness, the consumer impact of AI-powered brand communications is conditional upon perceptions of brand messaging as being trustworthy and human-centered, socially meaningful. The study constructs a consumer-behavior framework of the impact of human-centric AI-driven social media intelligence on trust in AI-based brand content, e-WOM, pur-chase intention and perceived sustainability. Situated in the fields of digital marketing, social media intelligence and behavioral consumer analytics, this study aims to investigate a quantitative survey conducted among Indonesian social media users who have been exposed to AI-assisted or AI-generated brand communication. Data was analyzed with PLS-SEM with trust being treated as the inner psychological mechanism and e-WOM as the outer social amplification mechanism transferring AI-enabled marketing to purchase intention and perception of Sustainability. Moving beyond technological adoption, this research on AI marketing highlights customer intelligence, consumer trust construction, online recommendations and responsible digital value creation. In practice, the framework guides firms in designing AI-enabled social media strategies that are persuasive, credible, customer- and sustainability-oriented.
Authors - Chaithra G, Ambika P R, Manjunath R, Shivashankar, Niranjan R, Sowmya Naik P Abstract - Gross Domestic Product (GDP) forecasting using traditional Econometric models is indispensable for evidence-based decision-making. However, these models are often limited in their ability to handle linear relationships and adapt to high-dimensional data. This paper introduces EcoVision, an open-source web-based forecasting platform that incorporates AI and machine learning to accurately predict GDP and other associated socio-economic variables using the Gap minder dataset (175 countries, 1998-2018). Four machine learning models were used: Support Vector Machine Regression, Polynomial Regression, Decision Tree Regression, and Random Forest Regression. These were built using Python and the Flask/Scikit-learn stacks. Models were evaluated using Average Absolute Error, Squared Error, and R² values. Results show that the Decision Tree Regression model has a perfect fit (R² = 1.0, AAE = 0, SE = 0), making it the best model compared to the other models. The web interface is built using pure HTML5/CSS3/Chart.js. The integrated "Gemini API Module" enables the automatic generation of easily understandable policy summaries, thus allowing for faster extraction of insights. Results from testing the system on 3532 clean data records proved that the system is accurate in forecasting ≥ 85%, Artificial Intelligence summary relevance ≥ 80%, and export success 100%, making it a potential decision-support system for economists, researchers, and governments.
Authors - R Suganya, S Priya, Sheeja Pon Chakravarthy,Pragadheesh Thirumal M Abstract - This paper presents a real-time intelligent surveillance system designed to detect weapons and violent activities using deep learning techniques [2]. The system integrates the YOLOv7 object detection model [7] for weapon recognition and a CNN-based violence detection module for behavioral analysis. Real-time video streams from CCTV cameras are processed to identify potential threats, and alerts are transmitted via MQTT for immediate notification. Experimental evaluation demonstrates that the YOLOv7 model achieves a mean Average Precision ([email protected]) of 55.3% for weapon detection, while the CNN model [11] attains 96% accuracy in classifying violent actions. The system operates at an average speed of 25–30 frames per second with low latency, confirming its feasibility for live surveillance applications. The proposed architecture enhances public safety by providing automated, accurate, and real-time monitoring capabilities.
Authors - Melati Budi Srikandi, Jonathan Jacob Paul Latupeirissa, Yolanda Masnita, Rizki Dewantara, Ni Made Prasiwi Bestari, Ayu Made Bianca Juarez Abstract - The shift from human gatekeepers to AI-driven algorithmic curation has fundamentally changed the concept of "Agenda-Setting" theory in the digital age. This change is significant because AI now influences public issue salience; however, there is a notable gap in public awareness. This study examines AI's role in social media information curation and its effects on public discourse and agenda setting. To do this, the research employs a systematic literature review guided by PRISMA principles, analyzing data collected from the Scopus database to identify current research trends. Moving from “handheld” to “automatic” curation results in more personalized interfaces that foster “filter bubbles” and “echo chambers,” according to the analysis. It demonstrates that understandings of “algorithmic news bias” are more influenced by user partisanship and ideological cues than purely technical causes. In conclusion, this suggests that media theories need to be refined to include automated gatekeeping as a core component. Algorithmic literacy serves as a filter against distortions, aiming to reduce disinformation and digital conflict within society. To address the bottlenecks in communication processes, future research and policy should focus on improving algorithmic literacy, given its undeniable influence over human decision-making.
Authors - Helmy Wijaya, Vallencia Ricca Widjaja, Fernand Jetshen Clevanno, Ichwan Masnadi Abstract - With rapid digitalization happening in the hospitality industry today, hotels are now able to interact more digitally with their guests using innovative customer service solutions. Aspects of technology adoption can be studied from the perspective of customer intelligence to gain behavioral insights about customers, which in turn can help hoteliers improve their user experience with such technologies and create a higher rate of adoption amongst customers. This research explores what factors affect hotel customers' intention to adopt online check-in technology by implementing the Technology Acceptance Model (TAM) with an exploratory factor analysis aimed at customer behavioral insights. Using quantitative explanatory research methods, data from 150 respondents in Jakarta was gathered through online questionnaires. Structural equation modeling was analyzed through Partial Least Squares SEM (PLS-SEM). Empirical results showed that PU and PEOU affected users' attitude toward using online check-in technology. The users' disposition exerted a considerable influence on their aspiration to utilize the digital check-in technology. The effect of PU and PEOU on intention was fully mediated by attitude, implying how affective evaluation by customers has an impact on customers' behavioral intention.
Authors - Bhavya Balakrishnan, Srinivasa HP Abstract - The massive deployment of heterogeneous, resource- con-strained and always-on devices underlying the Internet of Things (IoT) has introduced complex cybersecurity challenges. The rapid growth of the Internet of Things (IoT) due to the large-scale deployment of heterogeneous, resource-constrained and always-on devices has resulted in complex cybersecurity challenges. The physical and digital components in the IoT systems are tightly bound which increases the attack sur-face and makes them highly prone to threats of malware infections, data theft, unauthorized access and distributed denial of service. Traditional security mechanisms and rule-based intrusion detection systems cannot manage the dynamic, large-volume and evolving IoT traffic. The solutions provided by machine learning have been widely concerned due to its capability of learning data patterns and finding abnormal and malicious activities. However, existing machine learning models have serious constraints such as lack of labelled information, extreme class imbalance, and inability to generalize to new and never-seen attacks. In recent years, Generative Adversarial Networks (GANs) have emerged as a promising paradigm to improve the cybersecurity of IoT through artificial generation of realistic synthetic data, adversarial sample enhancement, alleviating data imbalance and modelling adversarial attack-defense dynamics. GAN based models have showed great gains in intrusion detection, anomaly detection and malware analysis in the IoT networks . However, modern studies are still divided on this issue due to variations in GAN architectures, datasets, evaluation procedures, and experimental procedures. In addition, most of the researches have been more concentrated on offline benchmark databases, with less focus on checking through realistic IoT testbeds, which could be more precise in capturing the actual deployment conditions.
Authors - Ferry Setyadi Atmadja, Sabo Hermawan, Eka Dewi Utari, Suciati Putri Nurjanah, Siti Dwi Hastuti Abstract - The exorbitant costs associated with professional Content Management Systems (CMS) have precipitated a severe theory to practice gap in digital archive education. This infrastructural barrier disproportionately disadvantages institutions with constrained budgets, fundamentally threatening the inclusive education mandates of Sustainable Development Goal (SDG) 4. To bridge this ped-agogical divide, this study developed and validated a zero-license educational framework utilizing Microsoft Excel's Visual Basic for Applications (VBA) to simulate a professional electronic records environment. Employing an R&D methodology (ADDIE model) with a cohort of 40 undergraduate students, the proposed framework circumvented hardware and financial constraints by operating offline on low-specification devices. Results indicated high expert validation (4.35/5.0) and a statistically significant enhancement in students' practical archival skills, evidenced by a moderate to high Normalized Gain (N-Gain) of 0.61. Furthermore, the system demonstrated exceptional usability with a System Usability Scale (SUS) score of 76.5. These findings provide empirical evidence that strategic, low-cost technological interventions can effectively democratize digital archive learning, offering a highly scalable solution for marginalized educational ecosystems in developing regions.
Authors - Rayyan Naufal Anandito, Muhammad Fedylopa Ginting, Trias Septyoari Putranto Abstract - The rise of Automated Biometric Boarding Systems (ABBS) for public transportation, driven by the potential to enrich convenience while integrating artificial intelligence into their activities has not been without the desire among policymakers and business leaders to get a better grasp on how biometry could be integrated in mandatory adoption contexts. Abstract This study aims to investigate passenger acceptance and continuance intention of AI-based face recognition boarding system in PT Kereta Api Indonesia (KAI) Gambir Railway Station 2023. Based on an integrated framework of Technology Acceptance Model (TAM) and Expectation-Confirmation Model (ECM), complemented with Trust and Perceived Privacy Risk, this study explores the pathways through which affective factors and institutional factors influence long-term behavioral intentions in a compulsory acceptance context. Data from cross-sectional, quantitative. 150 purposively sampled passengers were analyzed by PLS-SEM using SmartPLS 4.0. This is the first time that these findings challenge many of the assumptions about technology adoption and provide relevant policy recommendations for transport authorities based on a framework for AI governance aligned with Indonesia's Personal Data Protection Law (UU No. 27/2022).
Authors - Ridhi Sharma, Ashok Kumar Abstract - This manuscript discovers the role of information theoretic measures for feature selection while dealing with high dimensional data sets. The study uses entropy, mutual information and divergence measures to address the issues of classification and high computational complexity of real data set which is affect by redundant and irrelevant features, by analyzing the dependency patterns and feature relevance in complex data set. Under different data conditions, the proposed approach for feature selection, in comparison to traditional methods, handles the non-linear relationships and noisy attributes effectively in terms of relevance, classification and interpretation. In-formation theoretic methods provide more precise feature selection and pattern identification results in the data sets. Despite the challenges of computational cost and scalability, the study shows that information theoretic measures can perform better in feature selection and decision making of the data mining.
Authors - Nishi Doshi, Shrey Shah Abstract - Hospitals run more machine learning on GPUs while the carbon footprint of grid electricity rises and falls through the day. Using a computer simulation, we compare 13 scheduling rules on mixed GPU hardware, with synthetic patient-style jobs, urgency tiers, and time-ofday carbon traces. We do not study patient outcomes; every percentage we report is a simulator queue number, not a clinical finding. We ask whether running non-urgent jobs overnight is almost as good as a richer rule that mixes urgency and carbon (CUCA at weight 0.45, written CUCA 0.45). The comparison keeps carbon reduction secondary to clinical priority and deadline compliance, so each policy is judged on both average kg CO2e and missed-deadline behavior. CarbonGreedy and CarbonShift are carbon-first stress tests that demonstrate how poorly wrong vendor presets can disrupt clinical priorities, and are not meant for production. Numbers are averages over many test settings, with wide run-to-run spread and no statistical adjustment, so headline ratios are exploratory. On an eight-GPU baseline, the overnight rule closes about 78% of the carbon gap between urgency-only and CUCA 0.45 while missing fewer urgent deadlines than either. CarbonShift lets about 46% of the most urgent jobs miss their deadline; this is simulated queueing, not bedside harm. At 48 jobs per hour, the carbon footprints almost tie, yet the overnight rule still misses fewer urgent deadlines. A geography test, where regions share one daily carbon shape with only timezone shifts, trims under one percentage point of average carbon; a twelve-hour routine window saves a little carbon for CUCA 0.45 but raises overall missed deadlines. Overnight batching stays competitive on average modelled carbon; carbon-only rules belong only in stress tests.
Authors - Abhijit Dnyaneshwar Jadhav, Prashant G. Ahire, Madhuri Hiwale Abstract - In vitro fertilization (IVF) is currently one of the most powerful assisted reproductive technologies for infertility treatment. However, the embryo selection process still represents a bottleneck that greatly influences the rates of implantation and live birth. Traditional methods of embryo evaluation involve embryo morphology grading. But this approach suffers from subjectivity, variability, and heavily depends on the skill and experience of the embryologist. To go beyond the limitations of human assessment, the latest improvements in artificial intelligence (AI), machine learning (ML), and deep learning (DL) have made possible the automated embryo evaluation using pictures, time-lapse morphokinetics, and clinical data. This paper reviews comprehensively the currently available AI-enabled IVF systems while also first introducing the conventional embryo assessment and later presenting the most sophisticated multimodal deep learning frameworks. The paper also discusses some of the major outstanding issues such as the poor performance of models on new datasets, the lack of the shared and agreed upon benchmarks, and the limited explainability of the models. We have also developed a Multimodal Explainable Artificial Intelligence Frame-work for IVF (MEAIF-IVF) to fill in these gaps in which image of the embryo, time-lapse video of the embryo, and clinical patient information are all combined into one deep learning model. This system uses convolutional neural networks and vision transformers for spatial feature extraction, recurrent neural networks for temporal modeling, and attention-based fusion for multimodal integration.
Authors - Peruru Gayathri, Rohini M, Anand R Nair Abstract - Cyber threats are getting more sophisticated and conventional security solutions are not keeping up with detecting cyber-attack. In this research, a hybrid detection and prediction system for TTP (Tactics, Techniques and Procedures) based on deep learning and graph-based is presented. The planned study is based on an analysis of data originating from cyber security systems at large scale, which can be used to detect attack patterns and correlations of attacks. Host logs and threat intelligence data are trained using deep learning models to detect discriminative features, while graph-based models are used to model the structural relationships between users, systems, and attack patterns. Combined these techniques will result in more complex attacks and lateral movement being easier to detect. It also assumes probable attack methods to move to the next level, so that it can predict the attacks and take proactive actions to mitigate attacks in the future. The entire predictive and graph based solution enhances threat visibility and threat response speed, while boosting threat detection accuracy. The system enables the detection of the APTs and real time monitoring them by the Cyber Security analysts. The experimental results show that the highest accurate transformer is able to achieve 95% classification accuracy, and the graph neural network is demonstrated to achieve 78.26% accuracy for predicting next technique. The framework has been shown end-to-end, with the intent of showing it can be utilized as an extra layer of Intelligence on the enterprise security side, with Splunk.
Authors - Anis A, Kasi B Anand, Adithyan B, Navaneeth A Nayan, Durgalashmi CV Abstract - The advent of the prospects of using startups especially tech-based having unprecedented significance and continuous severity with regard to Sustainable Development Goals (SDGs) has increased the understanding of how important it is to understand students' knowledge and perceptions in this field. This study is designed for survey to understand and measure students' depths of knowledge regarding startup companies, the SDGs, as well as potentially precautionary attitudes towards startups in achieving sustainable development by 2030. The robust quantitative research design, were a well-designed and systematically developed questionnaire was employed to identify a systematic collection of questionnaires from students who are studying in the different higher educational institutions by incorporating cross-sectional survey methodology data collection technique. In general, the results show that students agree that startups are good for economy, society as a whole and even environment. However, research also mentions that there is low perception about startup doing Sustainable development work along-with moderate awareness and moderate belief in strong government support. These parts together suggest deep necessary work by the policymakers, educators and other stakeholders to raise the level of awareness and support. Furthermore, the study demonstrates the need for systematic integration of sustainability and entrepreneurship education to enhance students' knowledge on sustainability issues as well as their involvement in sustainable development-oriented tasks. The findings of this study offer new and practical lessons to policymakers, educators and researchers facing the continuing challenge of building and reshaping startup ecosystems that reflect or foster their successful fulfilment of sustainable development goals.
Authors - Chethana R.M. and Dr S.P. Manikandan Abstract - The rapid evolution of cyber threats has intensified risks to organisational security, necessitating intelligent, data-driven approaches to threat assessment and mitigation. This study presents a comprehensive analysis of the evolving cyber threat landscape and its impact on organizational security using a dataset of 1,200 cybersecurity incidents reported across major sectors in India from 2019 to 2024. The dataset includes diverse incident categories such as phishing, ransomware, data breaches, online fraud, identity theft, and hacking, along with associated financial losses, geographic distribution, and affected organizational domains. To investigate threat patterns and predict incident behavior, three machine learning models, Support Vector Machine (SVM), Random Forest (RF), and Logistic Regression (LR) were employed for classification and regression tasks. Experimental results reveal significant challenges posed by class imbalance and feature complexity, leading to relatively low classification accuracies, with Random Forest marginally outperforming other models. Regression analysis for predicting financial losses also demonstrated limited explanatory power, indicating the influence of latent factors beyond the available attributes. Despite these constraints, the study identifies important sector-specific vulnerability patterns, highlighting significant financial impacts across healthcare, financial services, and government. The findings emphasize that conventional machine learning models alone may be insufficient for capturing the highly dynamic and nonlinear nature of cyber threats, underscoring the need for advanced threat intelligence frameworks, richer datasets, and adaptive security analytics. This research contributes empirical insights into cyber risk modeling and offers practical implications for policymakers and organizations seeking evidence-based cybersecurity strategies.
Authors - Xamdamov Utkir Raxmatillaevich, Elov Botir Boltayevich, Alavutdinova Nadira Ganiyevna, Malika Suyunova Odil qizi, Sharipov Soxib Salimovich , Narimova Gulnora Abdumanonovna Abstract - In this article, the architecture of an information system for sentiment analysis of Uzbek-language texts and its key components are examined from both scientific and practical perspectives. The system is based on a multi-layered and microservice architecture, consisting of a user interface (front-end) and a server (back-end) that provides services through a REST API. The back-end components, implemented via a Flask-based RESTful API server, carry out the business logic and sentiment classification. Deep learning models, especially transformer-based architectures (BERT, XLM-RoBERTa), were utilized for analyzing Uzbek texts and demonstrated effective results. The system ensures security, provides integration capabilities, and offers a user-friendly interface to enhance user experience. The modular architecture of the system allows broad scalability and integration with various platforms. As a result of scientific and practical experiments, the system achieved high accuracy (90%) and proved effective for real-time sentiment analysis tasks.
Authors - Sunandita Adhikary, Dipanwita Chakrabarty, Arunangshu Giri, Shamba Chatterjee, Dibyendu Rath, Soumya Kanti Dhara, Solanki Pattanayak, Samik Bagchi Abstract - The study evaluates how shopping engagement gets influenced by mobile apps in digital retail platform. In e-commerce platform it is important to understand user preference in the context of customization, quality of information, usability and interactivity. The present study investigates how these contextual parameters play a pivotal role in shaping users’ emotional and cognitive reactions. These reactions subsequently influence user engagement and purchase-related decisions. The study has proposed a structured framework to identify the antecedents’ influence on shopping engagement and how it shapes user satisfaction. The findings of the study shows that mobile app plays a crucial role in engaging user in digital retail platform and consequently users’ shopping engagement influence their choice satisfaction. The study has notable contribution for marketers and mobile app developers so that they can enhance users’ satisfaction and can achieve competitive advantage. The study has enriched existing literature as well by extending expectation confirmation theory in the context of shopping engagement through mobile application in digital retail platform.
Authors - Maria Genoveva Moreira Santos, Eric Geovanny Cedeno Zambrano Abstract - Document management in public institutions constitutes a strategic element for improving administrative efficiency and strengthening decision making. In this context, the present study analyzes the implementation of a centralized repository in Ecuador’s Water Secretariat, aimed at the use of business intelligence tools and platforms to optimize access to, control of, and utilization of institutional information. The study was conducted using a quantitative methodology, supported by interviews applied across different departments of the institution, in order to identify needs, limitations, and practices related to records and process management. The results revealed a low adoption of specialized technological solutions and a limited appreciation of their strategic potential within the public sector. A total of 83.7% of the results supported the need to establish clear regulations for document management. It is concluded that the integration of document management systems with business intelligence platforms promotes the generation of timely information, institutional monitoring, and evidence based decision making.
Authors - Eka Dewi Utari, Darma Rika Swaramarinda, Maulana Amirul Adha, Triesninda Pahlevi, Yuliansyah, Dewi Nurmalasari, Agung Kresnamurti Rivai P, Ferry Setyadi Atmadja, Fauzan Fadlullah, Alifah Kusumaningrum, Sabo Hermawan, Renata Rachel Abstract - Academic websites have emerged as critical intelligent digital infrastructures for delivering institutional information and services in higher education. However, existing evaluation frameworks often capture either technical quality dimensions or subjective user experience in isolation. This study proposes and empirically validates an integrated evaluation model combining WebQual 4.0 with the Ease of Use construct from the End User Computing Satisfaction (EUCS) model, applied to the official website of the Faculty of Economics and Business, Universitas Negeri Jakarta (www.feb.unj.ac.id). The integration is motivated by the growing imperative to align academic web services with intelligent service design principles encompassing data-driven content governance, responsive interaction channels, and user centred personalization as foundations for future AI augmented academic portals. A quantitative descriptive design collected data from 124 respondents (93.5% students; 6.5% lecturers) via a 17-item validated questionnaire across four dimensions: Usability, Information Quality, Interaction Quality, and Ease of Use. Multiple linear regression (IBM SPSS 23) revealed that Information Quality (β = 0.419, p < 0.001) and Interaction Quality (β = 0.260, p = 0.002) exerted statistically significant partial effects on user satisfaction, whereas Usability and Content did not reach partial significance. Collectively, the four dimensions explained 70.8% of satisfaction variance (R² = 0.708; F = 72.074; p < 0.001). Bibliometric keyword-network analysis contextualises the study within the broader digital-services literature. The integrated WebQual–EUCS model offers a replicable diagnostic tool for higher education institutions seeking to align web services with intelligent user expectations.
Authors - Anggita Sharon Simanjuntak, Eva Nurhazizah Abstract - This study investigates the impact of cognitive perception on destina tion trust and intention to visit, while examining the moderating role of social media influencers at Taman Impian Jaya Ancol, an urban tourism destination in Indonesia. Utilizing a quantitative approach, data were collected from 385 re spondents via purposive sampling and analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM) via SmartPLS. The results reveal that cognitive perception significantly enhances both destination trust and intention to visit. Similarly, destination trust and social media influencers exhibit a signif icant positive effect on visit intention and destination trust, respectively. How ever, social media influencers do not significantly moderate the cognitive per ception-destination trust relationship. Ultimately, these findings highlight the ne cessity of cultivating positive perceptions and trust, offering strategic insights for destination managers to optimize social media marketing.
Professor, National Institute of Technology Raipur, India.
Narendra D. Londhe is presently working as Associate Professor in the Department of Electrical Engineering of National Institute of Technology Raipur, Chhattisgarh INDIA. He completed his B.E. from Amravati University in 2000 followed by M.Tech and Ph.D. from Indian Institute of Technology... Read More →
Authors - Max Angelo D. Perin, Lenie B. Maligmat, Darrel A. Cardana, Renante S. Digamon, Joan Mae G. Lagumbay, Cecilia T. Gumanoy Abstract - The Quality Assurance Office of a Philippine state university campus conducts 7S evaluations across all offices each semester, producing numeric scores and written evaluator comments. Consolidating the narrative comments has depended on manual review, which is time-consuming across more than a hundred offices per cycle. This paper describes a two-phase AI-assisted analytics pipeline. Phase 1 retrieves audit records from a MySQL database via a stored procedure, formats them with a Python ETL script, and submits them to Grok (xAI) to draft scorecards and action items; evaluators then review the drafts be-fore consolidation into the official PDF report. Phase 2 parses the validated PDF with Python to extract structured fields and compute descriptive statistics, office rankings, a priority index, and TF-IDF text clustering. Applied to the November 2025–January 2026 cycle (112 offices; 107 scored, 5 with no submission), most units cluster in the moderate-to-great compliance range while a meaningful minority fall below threshold. Among the top 25 priority offices, Standardize (20/25) and Safety (19/25) are the most frequently flagged dimensions. The pipe-line shows that AI assistance structured around human review can accelerate QA consolidation while preserving evaluator accountability.
Authors - Nur Fajrina, Felina C. Young, Rosita Widya Putri Abstract - This study investigates the relationships among Stakeholder Integration (STI), Digital Transformation Capability (DTC), Absorptive Capacity (AEC), and Sustainable Supply Performance (SSP) within a knowledge-intensive supply chain context. Employing a quantitative methodology alongside Partial Least Squares Structural Equation Modeling (PLS-SEM), data were collected from 262 respondents involved in strategic and operational functions. The results reveal that stakeholder integration significantly enhances digital transformation capability, thereby strengthening absorptive capacity. Both digital transformation capability and absorptive capacity have direct positive effects on sustainable sup-ply performance. However, stakeholder integration does not directly influence sustainable supply performance. Instead, its effect becomes significant only when mediated by absorptive capacity, indicating that internal knowledge assimilation and utilization mechanisms are essential for translating collaborative efforts into sustainability outcomes. The results highlight the critical role of dynamic capabilities in accomplishing sustainable supply performance, particularly in environments characterized by digital transformation and stakeholder complexity. The study contributes theoretically by integrating stakeholder theory and dynamic capability perspectives, emphasizing absorptive capacity as a key mediating mechanism. The results suggest that firms should complement external stakeholder collaboration with investments in digital infrastructure and organizational learning systems to enhance long-term sustainability performance.
Authors - Ayush Ghumare, Reena S. Satpute Abstract - Mobile application development has evolved rapidly with the emergence of advanced technologies such as 5G connectivity, Artificial Intelligence (AI), Machine Learning (ML), and Mobile Edge Computing (MEC). These technologies are transforming the mobile ecosystem by enabling the development of intelligent, data-driven applications and accelerating development cycles. Mod-ern mobile applications are expected to provide real-time services, personalized user experiences, and seamless connectivity, which has significantly increased the complexity of mobile application design and implementation. It is resulting into many challenges. One of the major challenges in mobile application development is the inherent limitation of mobile devices, including restricted pro-cessing power, limited memory capacity, and battery constraints. Developers must optimize application performance while ensuring energy efficiency to pre-vent excessive battery consumption and degraded user experience. Additionally, the increasing reliance on third-party libraries and analytics tools may introduce security vulnerabilities, creating potential security gaps within applications. These risks are often intensified by the lack of specialized security expertise within development teams, raising concerns related to data privacy, application security, and software supply chain vulnerabilities. Another challenge is platform fragmentation, particularly within the Android ecosystem, where diverse devices, operating system versions, and hardware configurations complicate compatibility and performance optimization. This diversity increases testing complexity and development costs. Furthermore, integrating AI and ML models into mobile ap-plications requires careful decisions regarding cloud-based versus on-device pro-cessing. Therefore, developers must balance scalability, performance, security, and energy efficiency when designing modern mobile applications. This study presents systematic literature evaluation methodology, comparative analysis of native and cross-platform paradigms, software supply chain security frameworks, measurable energy optimization strategies, and practical industry case studies from healthcare, fintech, and mobile commerce sectors.
Authors - Murali Mohan Reddy Seelam, VyshnaviThanneeru, Ajay Kumar Reddy Vemireddy, Srilatha Kudumula Abstract - This paper shows a new approach to implement the Agentforce-NS framework to provide zero touch salesforce deployment pipelines by integrating it with the Neuro Symbolic AI Agents. Even though the complete salesforce deployment pipelines have been automated end to end, it has been very difficult to achieve zero touch deployments due to its nature of the handling of metadata due to the interdependency of the components within the salesforce. The regular pipeline processes still heavily depend on the manual intervention to resolve the merge conflicts, resolve the dependency errors, working on the roll back deployments and following the compliances. The architecture we are proposing will solve all these problems by integrating the adaptive and predictive capabilities of the neural networks with rule based, transparent precision of the symbolic reasoning. The proposed Agentforce architecture will have five agents that will collaborate and will execute the deployments without any human intervention. These five agents are used to learn the deployment strategies, roll back planning, analyzing the metadata, autonomous execution and verification of the governance. After many tests in the enterprise level environments, we see that it is resolving so many blockers, issues and increasing the deployment success rate, improving the governance, and reducing the meantime to recover. By covering the technical gap between logical interface and the deep learning, the Agentforce-NS represents a break through advancement to have the fully automated, autonomous and auditable salesforce devops pipelines.
Authors - Shreya S. Partake, Reena S. Satpute Abstract - Large Language Models (LLMs) have demonstrated remarkable capabilities in natural language processing, achieving human-level performance on many semantic and syntactic benchmarks. However, their competence in pragmatics—the study of how context shapes meaning—remains a critical and underexamined frontier. This paper presents a unified analysis of the “pragmatic gap” in LLMs, arguing that it stems from a fundamental distinction between the co-textual statistical patterns LLMs are trained on and the contextual world knowledge humans use for inference. We first establish a theoretical baseline by reviewing foundational linguistic concepts, including Grice’s maxims, implicature, presupposition, speech acts, and deixis. We then systematically evaluate LLM performance, contrasting successes in pattern-rich tasks like coreference resolution with systemic failures in tasks requiring novel inference, such as non-conventionalized indirect speech acts and irony. We analyze the development of new evaluation tools, particularly the Pragmatics Understanding Benchmark (PUB), which quantifies the persistent gap between model and human performance. Subsequently, we synthesize emerging technical solutions, including “thought-based” fine-tuning and the injection of Gricean principles into Retrieval-Augmented Generation (RAG) frameworks. Finally, we dissect the profound cognitive and philosophical implications of this gap, critically examining the debates on the Symbol Grounding Problem and Theory of Mind (ToM). We conclude that while LLMs can pass “literal” ToM tests, they fail “functional” ToM, revealing them to be sophisticated co-text manipulators rather than context-aware agents. We propose that future progress lies in developing a “machine pragmatics” based on probabilistic models rather than flawed anthropomorphic imitation.
Authors - Ayushi Chapate, Reena S. Satpute Abstract - Natural language helps us to interact with the computer through human language. This article investigates how Natural Language Processing (NLP) can enhance our understanding of social media changes. To its audience, social media provides a large - arguably unlimited - and otherwise untapped linguistic re-source, revealing information about government behavior, civic participation, in-dividual mental well-being, and consumption behavior, among many other things. Using machine learning analytical methods such as sentiment analysis, topic modeling, stance detection, and misinformation tracking, researchers can begin to study the social, psychological, and economic implications of web-based inter-action. In terms of civic and political implications, to analyze user-generated con-tent, discourse networks, and hashtags using NLP applications can produced new insights into online mobilization and collective action. For example, researchers studying the political movement’s #MeToo and #BlackLivesMatter, based on analysis of Twitter data, have employed topic modeling techniques to reveal their influence and significance in innovative ways. From a psychological perspective, NLP methods make it possible to examine prevalent mental health indicators across separated populations, through the analysis of emotional tone, pronoun use, and distress markers. In studies conducted between 2020–2025, the application of BERT based embedding models were found to detect online indicators of depression, anxiety, and social comparison leverage's based on word meaning. Further, understanding the depth of these psychological consequences remains nebulous and limited to a range of social categories in the digital landscape, similar to previous notions of 'self-checking' across the digital commons exploring citizen engagement.
Authors - Jes Maries M. Mendez, Max Angelo D. Perin, Joan Mae G. Lagumbay, Mae S. Dagupan, Elizabeth A. Orapa, Marcelina S. Butlig Abstract - Educational tours are widely used in higher education to connect class-room learning with real settings, yet evaluations often stop at overall ratings that do not explain why students endorse a tour or which delivery issues weaken the experience. This study applies a student experience intelligence workflow that integrates survey analytics with offline text mining to produce planning-relevant evidence. A survey of 156 students captured demographics, three 10-item Likert constructs—motivation, perceived effectiveness, and problems encountered (4-point scale)—a recommendation rating, and open-ended comments. Responses were cleaned through category standardization and rule-based numeric conversion. Internal consistency was good for motivation (α = 0.877) and excellent for effectiveness (α = 0.960) and problems (α = 0.958). Learning beyond classroom instruction (M = 3.71) and interest in tour inclusions (M = 3.68) led motivation; creative learning (M = 3.67), resourcefulness (M = 3.66), and social skills (M = 3.65) led effectiveness; tour expense (M = 3.21) and short time per attraction (M = 2.60) led problems. 73.1% gave the top recommendation. Recommendation correlated positively with motivation (ρ = 0.317, p < 0.001) and effectiveness (ρ = 0.328, p < 0.001); a binary logistic model showed perceived effectiveness as the strongest predictor of the top recommendation category. Open-ended comments (171 entries) were summarized through TF–IDF with K-Means clustering (k = 6) and complemented with a VADER polarity pass on 155 meaningful entries (68.4% positive, 21.9% neutral, 9.7% negative; mean compound = +0.365). The combined evidence points to improvements that preserve educational value while addressing cost and pacing, and shows that the workflow is portable to other programs and experiential learning activities.
Authors - Kent Cyryl A. Campit, Christian Kelvin Gonzales Abstract - This study explored a data-driven approach to evaluating citizen feedback within the Provincial Government of La Union (PGLU) by integrating quantitative and qualitative analytical techniques. Traditional feedback systems in government offices often rely on averages and summary reports, limiting the ability to capture deeper citizen experiences and concerns. To address this gap, the research transformed paper-based feedback forms into a structured digital dataset covering responses from 34 frontline offices and service units from July 2025 to January 2026. The study applied Customer Satisfaction Score (CSAT), Weighted Mean, and Range of Interval to measure and classify service performance levels. For qualitative analysis, Latent Dirichlet Allocation (LDA) was used to identify recurring themes in open-ended responses, while a dual-model sentiment analysis approach combining VADER and RoBERTa classified citizen feedback into positive, neutral, and negative sentiments. The analytical pro-cesses were implemented using Microsoft Excel, Google Sheets, and Python through Google Colaboratory. Findings revealed consistently high satisfaction ratings across offices, while qualitative analysis uncovered recurring themes related to service efficiency, staff assistance, facility conditions, and operational concerns. RoBERTa demonstrated better contextual understanding and achieved higher performance metrics compared to VADER. The study further developed an Observed Satisfaction Classification Framework to support evidence-based decision-making and service improvement. Ultimately, the re-search demonstrated how citizen feedback can be transformed into actionable governance insights that promote transparency, accountability, and continuous improvement in public service delivery, aligned with Sustainable Development Goal 16.
Authors - Megha Potdar, Andhe Dharani, Ch.Ram Mohan Reddy Abstract - Semiconductor fabrication processes suffer significant yield losses, often exceeding 20%, due to equipment anomalies in critical stages like plasma etching and lithography, where traditional Statistical Process Control fails to detect subtle, non-linear drifts in multivariate sensor data such as temperature, pressure, and gas flow. This paper proposes a novel hybrid AI framework combining Long Short-Term Memory Autoencoder for unsupervised reconstruction-based anomaly detection with Isolation Forest for robust outlier scoring and severity ranking, enabling real-time predictive maintenance and Remaining Useful Life estimation. The LSTM-AE compresses temporal sequences into a latent space and flags anomalies via elevated Mean Squared Error thresholds (>95th percentile), while Isolation Forest filters multivariate errors to minimize false positives. RUL prediction employs linear regression on error trends for proactive scheduling. Implemented in a Keras/TensorFlow MLOps pipeline with
Authors - Kamasani Vishnuvardhan Reddy, Anjan Babu G Abstract - Universities maintain extensive repositories of institutional knowledge, yet students struggle to extract accurate information from disparate sources such as PDF circulars, web portals, and notice boards. Rule-based chatbots handle only narrow query sets, while general-purpose large language models (LLMs) produce fluent but sometimes fabricated responses—a phenomenon termed hallucination. This paper presents the Intelligent Campus Assistant Chatbot for Sri Venkateswara University (SVU), employing a Retrieval-Augmented Generation (RAG) pipeline that grounds every response in verified institutional documents via dense semantic vector search and deterministic keyword retrieval fused through Reciprocal Rank Fusion (RRF). Evaluation on a 200-query benchmark yields 94.2% factual correctness, hallucination rates below 1%, mean latency of 0.8 s, and inter-rater agreement κ = 0.87 across English, Telugu, and Hindi.
Authors - Ronald S. Cordova, Rowena O. Sibayan, Hazel C. Tagalog Abstract - Digital marketing teams often struggle less with access to algorithms than with choosing the right one for a specific decision. This paper presents a comparative study on the selection of the three most suitable algorithms for two related digital marketing tasks: customer segmentation and promotion-response prediction. Based on the example of Oman's retail industry, a benchmark is established using first-party customer data, including recency, purchase frequency, monetary value, product-category behavior, campaign participation, website visits, and engagement ratio. For customer segmentation, the study focuses on Kmeans, DBSCAN, and Gaussian mixture model because they provide a practical balance of scalability, noise handling, and probabilistic customer-state representation. For promotion-response prediction, the selected models are logistic regression, random forest, and XGBoost because they offer a staged balance between transparency, nonlinear learning, and campaign-ranking performance. For benchmarking and explainability, the same preprocessing approach, leakage prevention, temporal splitting, tuning strategies, and metrics such as silhouette quality, stability, ROC-AUC, PR-AUC, Brier score, calibration, and top-decile lift are employed. Explainability is treated as a condition for adoption rather than an optional reporting activity.
Authors - Mike Philip T. Ramos, Andres R. Vicedo, Jocelyn O. Padallan, Jonardo R. Asor, Genemarck B. Catedrilla Abstract - This research aims to develop a model for plankton species classification by analyzing images utilizing a convolutional neural network or CNN to simplify the task of classifying plankton species. The use of CNN and other transfer learning models will be used to recognize different freshwater plankton species in order to identify the genus of plankton easily. There were several layers in the CNN architecture used in this study; (1) Layer 1 has convolutional data with 32 filters and 3x3 kernel with max pool of 2x2 kernel; (2) Layer 2 has convolutional data with 64 filters and 3x3 kernel with max pool of 2x2 kernel; and (3) Layer 3 has conventional data with 128 filters and 3x3 kernel with mas pool of 2x2 kernel. After the validation and training in terms of accuracy and loss for CNN and pre-trained models, it is observed that MobileNetv2 showed the highest positive scores with 0.99 in train accuracy, 0.93 in validation accuracy, 0.07 in train loss, and 0.12 in validation loss, which makes it more viable to be used in this study. CNN's capacity to extract characteristics from photos has shown to be highly effective at classifying images. Additionally, it has been determined that transfer learning strategies can aid CNN in enhancing its picture categorization capabilities. The use of pre-trained learning like MobileNetv2 with a small data set and image classification studies can be a greater help for identification than CNN, Convolutional Network, Rest- Net50 and EfficientNetB0
Authors - Wannakorn Phornprasert, Nisarat Onthong, Thapanapong Sararat, Wongpanya S. Nuankaew, Pratya Nuankaew Abstract - This study proposes an Educational Data Analytics approach to understanding students' digital behavior and academic achievement using Descriptive and Cognitive Analytics. Data were collected from 40 purposively selected students using questionnaires that covered general information, social media usage, sleep behavior, Kolb-based learning style, and GPA. Descriptive Analytics was applied to summarize frequencies, percentages, means, and key behavioral patterns, while Cognitive Analytics was used to interpret these patterns in relation to learning readiness, self-regulation, and academic achievement. The findings showed that students had an average GPA of 3.38, spent an average of 7.53 hours on social media per day, and most frequently used social media between 20:01 and 00:00. The most common bedtime was 01:00, and Concrete Experience was the dominant learning style. The results suggest that small-scale learner data can support understanding of digital behavior, sleep patterns, and academic achievement in Thai higher education.
Authors - Kuljira S. Nuankaew, Kaewpanya S. Nuankaew, Wongpanya S. Nuankaew, Keingkrai Buakeaw, Thapanapong Sararat, Pratya Nuankaew Abstract - This research presents the development of an Explainable Intelligent Document Recognition system and a decision support system for assessing tax deductions in Thailand. The system uses image processing and data extraction technologies to analyze photographic documents and PDF files. It incorporates image quality enhancement, text recognition, key information extraction, and tax condition assessment, along with a rationale to enhance transparency in decision-making. Experimental results demonstrate efficient and accurate data recognition and extraction, and the system can handle diverse document types. Furthermore, a web-based prototype evaluation by 30 users showed high satisfaction, particularly regarding understanding the results and explanations. However, the system exhibits limitations with low-quality and complex documents. This research highlights the potential for applying such technology to taxation and for future expansion to improve flexibility and efficiency.
Authors - Najera R. Umpar, Apolinar P. Datu, Minsoware S. Bacolod, Soraya R. Umpar, Darwin B. Reyes, Albert Lee A. Catibayan, Klifford L. Carlos, Francia F. Murao Abstract - The widespread use of artificial intelligence (AI) tools into the world of academia has brought about substantial changes to the way scholars write. This paper examines how faculty view AI tools for use in academic writing through their own experiences of using these tools. Also, it explored their capacity to produce research publications and the integrity of the research being produced. Using a qualitative research design, the study gathered data through semi-structured interviews with faculty selected purposively using AI enabled tools such as ChatGPT, Sci.ai and Grammarly) during their writing process to collect data. Thematic Analysis was utilized to identify common themes within faculty member's accounts of their experiences. The findings of the study indicate that faculty perceive AI tools as valuable to enhance the speed in which they complete writing tasks, and also to improve grammar usage while writing, and to assist in idea generation; however, there were concerns voiced about overusing AI tools, ethical concerns with using AI tools, and how AI tools affect a faculty member's ability to think critically and produce work that is original. Additionally, the digital literacy level of faculty members who participated in this study reflects their ability to be able to adopt and incorporate these technologies into their daily teaching and research activities; thus, varying levels of digital literacy influence how a faculty member adopts and incorporates these technologies into their academic productivity. The study underscores the need for clear institutional guidelines and capacity-building initiatives to ensure the responsible and effective use of AI in academic writing. By providing insights into faculty experiences, this research contributes to the growing discourse on AI integration in higher education and offers implications for policy development, pedagogical practices, and future research.
Authors - Pradeep Kumar, Balasubramanian, Dhivyalakshumi Abstract - This paper analyzes the role of archetypal storytelling as an ethical brand meaning construction strategic tool in cause-related advertising as applied longitudinally to the Jaago Re campaign created by Tata Tea. Jaago Re is a cause marketing effort spanning more than 10 years and dealing with civic engagement, gender equity, community health and climate accountability. Based on the theory of archetypal branding, the paper examines thirteen aired and online advertisements published between 2008 and 2023 to learn how archetypes are utilized and redefined based on the changing socio-cultural and ethical issues. Based on the principles of a qualitative content analysis, guided by the Archetypal Criticism framework and Cultural Branding theory, the research paper recognizes the primary and secondary archetypes and investigates their narrative and ideological roles. The results have shown that archetypes that include the Hero, Sage, Caregiver, Everyman, and Outlaw, together with the Magician are well-planned layers that deploy civic actions, promote ethical contemplation, and maintain symbolic continuity in the campaign stages. The work proves that Jaago Re goes beyond episodic cause promotion, including the responsibility and social awareness as a part of the cultural identity of the brand. This study can be of use in the literature on ethical branding and responsible advertising because it links the progression of archetypal arrangements through time, providing a conceptual framework to build a sustained social and cultural value in the marketing communications.
Authors - Vishwa Kumaresh Abstract - A local supplier delay or demand shock in multi-echelon supply chains can make upstream orders volatile long before the full costs appear in planning dashboards. In this study, we propose an interpretable warning-to-action layer for supply-chain digital twins. This layer sits above the replenishment controller: it estimates disruption-regime risk from rolling demand, inventory, backlog, order, and lead-time telemetry, then maps that risk to bounded changes in responsiveness, safety stock, and order caps. We calibrate a gradient-boosted stump classifier that combines standard warning indicators, cross-echelon imbalance measures, and nonlinear stress descriptors. A small mode table converts the resulting probability into five auditable replenishment modes. This method is tested on twelve disruption scenarios grouped into six mechanism classes, using ten baselines and an untouched lockbox of 576 observations. The proposed policy reduces aggregate system expenditure by 15.2% and cross-echelon volatility (bullwhip) by 44.5%, relative to a linear guard that uses the same broad action family. The largest gains occur in lead-time disruptions and backlog cascades. Compound shocks demonstrate marginal performance gains, as existing linear guards effectively capture these dynamics within standard monitoring frameworks.
Authors - Wannakorn Phornprasert, Waraporn Phothirin, Thapanapong Sararat, Wongpanya S. Nuankaew, Pratya Nuankaew Abstract - This study uses Learning Analytics to assess university students’ eye health risks based on social media usage data, focusing on descriptive and diagnostic analyses. Data collected from 44 undergraduates via a self-reported questionnaire with 82 key questions covered general details, social media habits, device and screen environments, symptoms of Computer Vision Syndrome, and Felder–Silverman learning styles. The descriptive analysis revealed Instagram as the most popular platform, frequent nighttime use after 20:00, and many students spend over six hours daily on social media. While most respondents were categorized as low risk, symptoms such as watery eyes, eye pain, light sensitivity, and neck pain were commonly reported. The diagnostic analysis linked risky sitting postures, looking below eye level, prolonged daily usage, and nighttime social media activity to increased eye health risks. These findings support initiatives for digital well-being and learning support in higher education.
Authors - Mariel Leo T. Violeta Abstract - The increasing incidence of academic credential fraud, inefficient verification procedures, and reliance on centralized record management systems present significant challenges for higher education institutions. This study proposes and evaluates a blockchain-based academic credential issuance and verification platform using Hyperledger Fabric to improve the security, authenticity, and efficiency of academic credential management. The platform enables university registrars to issue digital academic credentials, allows students to securely access and share academic records, and provides employers and external entities with a reliable credential verification mechanism. To ensure data integrity while maintaining scalability and privacy, the framework integrates blockchain-based cryptographic hashing with off-chain cloud storage. A quantitative descriptive research design was employed using the Technology Acceptance Model (TAM) as the theoretical framework. Data were collected from 40 registrar personnel at the Polytechnic University of the Philippines through a structured survey instrument measuring Perceived Usefulness and Perceived Ease of Use. Findings revealed that respondents strongly agreed that the platform improves security, credential verification, operational efficiency, accessibility, and flexibility. The results demonstrate that Hyperledger Fabric can provide a secure, tamper-resistant, and efficient infrastructure for managing academic credentials in higher education institutions. The study contributes to the growing adoption of blockchain technology in education by presenting a practical and institution-oriented framework for secure and verifiable digital credential management.
Authors - Bhagyalakshmi S Pai, Jeevanand E S, Radhika P.C, Krupa B Nair, Sreeja Radhakrishnan, Dhanalakshmi Menon Abstract - The present study attempts to empirically investigate how the customers’ awareness relates to the adoption of green banking initiatives of commercial banks in Kerala, India. The study employs data gathered from 540 customers of five banks (SBI, Canara, PNB, ICICI Bank, HDFC Bank, and Axis Bank) by using a structured questionnaire, and builds and validates the structural model for green banking adoption. Customer awareness is considered as a higher order construct which consists of Environmental Awareness and General Awareness. The analysis used descriptive statistics, reliability analysis, Confirmatory Factor Analysis (CFA), two-stage analysis of Structural Equation Modeling (SEM), and Z test and One-Way ANOVA test to determine awareness levels and differences in demographic data. The results show that there exists a high Awareness–Adoption Gap, that is, a superficial awareness of green banking, which is not yet accompanied by a conceptual understanding of it. The study also reveals that adoption of e-banking is mainly for convenience and that practice in key life-stages and occupations have a strong bearing on adoption behaviour.
Authors - Wannakorn Phornprasert, Ratchanin Intham, Thapanapong Sararat, Wongpanya S. Nuankaew, Pratya Nuankaew Abstract - This study explored learning well-being and indicators of academic burnout associated with pseudo depression risk among university students at the University of Phayao. Data collection involved a general information questionnaire, an academic burnout assessment scale, and the DASS-21. Descriptive and diagnostic statistics were applied. Results indicated a moderate level of overall academic burnout, with academic fatigue scoring higher than academic withdrawal. Emotional risk assessment found that 50.0% of students showed mild to severe pseudo depression symptoms. Additionally, scores for academic fatigue, academic withdrawal, and overall burnout were positively linked to depression, anxiety, and stress. These results suggest that descriptive and diagnostic approaches can serve as initial tools for screening and promoting students' learning well-being in Thai higher education.
Authors - Pratya Nuankaew, Panisara Paksasuk, Thanapon Thiradathanapattaradecha, Thapanapong Sararat, Wongpanya S. Nuankaew Abstract - This study analyzes student behavioral data to understand factors influencing secondhand fashion purchases in the digital age. A survey was conducted with 40 University of Phayao students who are experienced in buying secondhand fashion items. Data analysis included descriptive statistics and diagnostic approaches to profile students, their purchasing habits, perceptions, and key factors. Results indicated that all participants had prior secondhand shopping experience, using both physical stores and online platforms as key channels. Product quality received the highest average score of 4.20, followed by a positive attitude toward second-hand fashion at 4.05, frugality at 4.00, and brand reputation and environmental responsibility at 3.85, with sustainable fashion close behind at 3.83. These findings suggest that students’ choices are influenced more by quality, value, personal attitudes, and sustainability awareness than by social media influencers alone. The research provides valuable insights for promoting sustainable fashion, designing platforms, and developing future predictive analytics.
Authors - Pratya Nuankaew, Duangjai Pongsawan, Supan Tongphet, Thapanapong Sararat, Wongpanya S. Nuankaew Abstract - This research aimed to examine the use of student-pet interaction data to enhance understanding of university students' mental well-being. Descriptive and diagnostic data analyses were conducted. The sample comprised 40 students. Data collection was conducted using questionnaires to collect baseline information, characteristics of interaction with pets, and evaluations with the CCAS, PSS-10, and ST-5 instruments. The analysis revealed that the majority of students experienced a high level of attachment and comfort with their pets, with an average CCAS score of 3.57. The average PSS-10 score was 20.48, indicating moderate stress levels, and the mean ST-5 score was 7.43. Diagnostic analysis suggested that the duration of contact with pets, pet type, living conditions, and pet ownership status were potentially associated with students' stress levels. These findings may serve as an initial guideline for developing monitoring and support programs to promote the mental well-being of university students.
Authors - Wannakorn Phornprasert, Papimon Novichai, Thapanapong Sararat, Wongpanya S. Nuankaew, Pratya Nuankaew Abstract - This investigation aimed to analyze the VARK learning style and ergonomic data to identify the risk of office syndrome among university students. A quantitative, cross-sectional approach was employed, utilizing questionnaire data from 40 students. The analysis used descriptive statistics to summarize general characteristics, learning styles, and risk levels, and diagnostic analyses to identify factors associated with office syndrome risk. The most prevalent learning styles identified were Read/Write (30.0%) and Kinesthetic (25.0%). Ergonomic assessments revealed that 42.5% of students were at high risk, while 35.0% were at moderate risk. Factors correlated with risk included excessive phone usage (exceeding 4 hours per day), inappropriate chair height, unsuitable armrests, incorrect screen positioning, and improper keyboard posture. These findings indicate that combining learning preferences with ergonomic data can serve as an initial screening tool for risk assessment and facilitate the development of learning environments tailored to students in the digital era.
Authors - Virgel William Afaga, Patrick Andrew Balang, Dana Wynnette Binwag, Emmanuel Paolo Bromeo, Mark John Bumacod, Carl Allan Calsiman, Juliana April Cendana, Roderick Makil,Dulthe Carlo Munar Jr. Abstract - Benguet Province, Cordillera Administrative Region, Philippines, is highly susceptible to landslides due to its rugged topography, complex geology, frequent typhoon tracks, and extensive mining and road construction. Existing hazard maps rely on static statistical methods and coarse rainfall averages that cannot capture the dynamic triggering conditions of individual storm events. This paper presents a dynamic landslide susceptibility model built on Random Forest (RF) and Extreme Gradient Boosting (XGBoost) trained on thirteen environmental conditioning factors across five domains (topographic: elevation, slope, aspect, distance to streams; geological: rock type, soil type; land cover: LULC, NDVI, NDWI; climatic/hydrological: mean annual rainfall, event rainfall, antecedent rainfall; anthropogenic: distance to roads) derived from high-resolution satellite imagery and event-specific rainfall records. Training used a balanced 16,158-sample dataset (50:50 landslide/non-landslide) from the MGB-CAR geohazard inventory, split 60:20:20 for training, validation, and testing. XGBoost outperformed RF on all metrics: AUC-ROC 0.8903, accuracy 81.78%, precision 81.87%, recall 81.62%, and F1 81.75%; the performance difference was statistically significant (McNemar's test: χ² = 6.22, p < .013). Spatial validation via the Seed Cell Area Index (SCAI) confirmed that High and Very High susceptibility classes captured 69.87% of inventoried landslides within only 36.3% of the provincial area. Expert review by four MGB-CAR geoscientists yielded Likert mean scores above 4.0 for conditioning factor appropriateness, inventory quality, and feature importance plausibility. A fully automated monthly update pipeline was deployed—completing the full cycle from remote-sensing data retrieval to web-map publication in approximately 31 minutes—demonstrating operational feasibility for dynamic hazard monitoring using open-source tools and free satellite data.
Authors - Adibhav Agrawal, Nikunj Parikh Abstract - This article is about how a new configuration of devices has been created for a compact, low-cost, real-time monitoring system for measuring electromagnetic fields on dairy farms. The Electromagnetic Field Monitoring System (EMFMS) is composed of an ESP32 micro-controller, MLX90393 three-axis magnetometer, TP4056 based boosting supply module, and a 0.96-inch OLED screen, which are all encased in a unique 3D printed PETG enclosure. The EMFMS can store and transmit wirelessly time-stamped activity levels of the earth’s magnetic field on three axes through MQTT protocol. The EMFMS was placed into three different areas of an operational dairy barn over 28 days where EMF levels of up to 17 times higher were observed between different areas, and a statistical finding was noted between EMF levels and lower levels of milk production (r = −0.61, p = 0.003) and higher levels of cortisol in serum (r = +0.44, p = 0.03) in Holstein-Friesian dairy cows. The findings of this pilot study demonstrate that this method of continuous measurement of electromagnetic fields for animals using IoT technology could serve as a more feasible and low-cost alternative to existing spot measurements.
Authors - Viraj Bhatt, Rajvi Bhimani, Bhupendra Fataniya, Dhaval Shah Abstract - Cross-border security remains a critical concern for global stability, particularly in jungle or forested terrains where soldiers face significant risks. Military camouflage is engineered to blend in with natural surroundings using advanced concealment techniques that match local textures and color patterns. Consequently, the identification of concealed threats is a challenging task where human observation is prone to error due to poor visibility and fatigue. Traditional surveillance methods often rely on optical sensors which may fail to efficiently detect modern military camouflage. To address this, an automated detection model was developed using the YOLOv8-Nano architecture and deployed on NVIDIA Jetson Nano hardware. The framework was validated using a 5- fold cross-validation strategy to ensure robust and reliable performance. Experimental results yielded a peak mean Average Precision (mAP) at an Intersection over Union (IoU) threshold of 0.5 of 0.955 and an average mAP of 94.8%. The model was further optimized into a TensorRT engine using FP16 quantization, achieving a final footprint of 5.9 MB. These results demonstrate that low-power, portable hardware can effectively perform real-time surveillance as an edge-AI system. This also results in minimizing risks to human lives and directly supporting the core mission of Sustainable Development Goal-16 (SDG-16).
Authors - Kapil Mohan, Ritu Chauhan, Harleen Kaur Abstract - ESG (Environment, Social and Governance) rating in today’s financial world is becoming a good indicator for investors in decision making and risk analysis. There has been stress on E and S in the recent past as Governments and Regulators stress these parameters and benefits to those who are working towards this improvement rating. The rating is a clear indicator of sustainability and promising business and thus is gaining popularity. The analytics firms have combined this indicator and have come up with this calculation using certain scientific and mathematical models from the published data and/or requested data that are provided exclusively to do this calculation for the indicator. These ratings are published annually by analytics firms like Sustain analytics and Bloomberg ESG data service for global but limited firms. This study’s focus is to fit financial data of firms on machine learning models and predict ESG rating with changing market fundamentals and firm’s business value indicators. The result can be com-pared to passed ratings, category averages, deviation and outliers which can benefit venture capitalists and investors to refine their investment strategies. The re-search also captures and compares this output and suggests the approach that best suits this problem by building an architecture that can update the model and can predict live data from the market.
Authors - Anuj Kothawade, Ishan Patra, Pravin Chavare, N. S. Shirude Abstract - The rapid digital transformation of higher education emphasizes the need for robust, data-driven platforms to monitor and enhance student learning. However, many institutions rely on closed, third-party learning management systems that restrict direct access to raw educational data and limit customized analytical capabilities. To address this gap, this paper proposes a scalable educational assessment and learning analytics platform that grants educators complete data sovereignty. Built on a modern stack of TypeScript, React and Tailwind CSS over an owned, directly accessible Firebase backend, the system enables secure, unhindered access for granular data mining. The platform monitors a range of college assessment activities, targeting quizzes and practical coding tests, and uses role-based authentication and custom data-fetching hooks to process student interactions into comprehensive performance metrics. A distinguishing feature is its integrated client-side PDF generation, which instantly produces detailed analytical score reports that serve a dual pedagogical purpose: empowering teachers with actionable insights to adapt instruction, while giving students personalized, self-reflective feedback for continuous improvement. Validated on a controlled pilot, the system achieved 95% overall accuracy, an 85% quiz-evaluation accuracy, a 28% improvement in student engagement, a 40% reduction in report-generation time, and a 92% system-usability score.
Authors - Aleta Fabregas, Nathanael Almazan, Jordan Garcia, Shaina Laman, Paolo Morato, Armin Coronado, Montaigne Molejon, Mariel Leo Violeta Abstract - The Philippines is frequently affected by tropical storms, typhoons, and flooding events that threaten communities located near major river systems. Accurate river level forecasting is essential for improving disaster preparedness and reducing flood-related risks. This study proposes RIVERCAST, a forecasting system that utilizes an Auto-Regressive Transformer with Kernel Principal Component Analysis (Kernel PCA) and Euclidean Kernel to predict Marikina River water levels across the Nangka, Sto. Niño, and Montalban monitoring stations. Meteorological, hydrological, and topographical datasets were collected from PAGASA, MMDA, DPWH, and OpenWeather API records from January 2012 to January 2023. Eighty percent of the collected records were allocated for training while the remaining twenty percent were utilized for testing. The pro-posed model was compared with the Transformer model developed by Xu et al. (2023) using rolling window testing and mean absolute error analysis. Results revealed that the proposed Auto-Regressive Transformer with Kernel PCA and Euclidean Kernel achieved an overall forecasting accuracy of 93.19%, outperforming the Bidirectional Transformer model, which obtained 92.57% accuracy. Findings further indicated that precipitation, rainfall intensity, and temperature significantly influenced forecasting performance, while humidity exhibited the least contribution. The developed model demonstrated reliable twelve-hour river level forecasting capability and may support flood preparedness and early warning initiatives within flood-prone communities along the Marikina River.
Authors - Syafira Aulia Azzahra, Christina Angelica Himawan, Brigitta Vellia, Indra Kusumawardhana Abstract - The hospitality industry is increasingly shaped by smart technology, integrated systems, and subscription-based service models that require consistent and personalized guest experiences. In Indonesia, particularly in the Greater Jakarta area, hotels are adopting Internet of Things-based devices, cloud-based property management systems, and data-driven service platforms to improve guest convenience and strengthen customer retention. This study examines the effects of smart technology devices and integrated systems on customer satisfaction in subscription-based hospitality, with service personalization positioned as a central mechanism in the guest experience. A quantitative cross-sectional survey was conducted with 400 hotel users in the Greater Jakarta area who had experience using smart technology in hotel services. The data were analyzed using Partial Least Squares Structural Equation Modeling. The findings show that smart technology devices and integrated systems positively influence customer satisfaction and service personalization. Service personalization also emerges as the strongest predictor of customer satisfaction, indicating that technology creates greater value when it enables relevant, adaptive, and individualized services. The study contributes to hospitality technology and customer intelligence literature by explaining how digital infrastructure and system integration support personalized subscription-based hotel experiences. Practically, the findings suggest that hotel managers should prioritize investment in interoperable systems, guest data integration, and personalization capabilities to improve satisfaction and sustain long-term customer relationships in technology-enabled hospitality services.
Authors - Aye Mya Mya Win, Ah Nge Htwe Abstract - In recent years, optical flow-based deep learning methods have pro vided evidence of impressive performance in recognizing human behavioral movements from video sequences, revealing high applicability for fall detection functions. This paper analyzes GMFlow-based architectures by experimenting with three different approaches that merge TCN, Attention, and CNN compo nents. These methods are GMFlow-TCN, GMFlow-TCN-Attention, and GMFlow-CNN-TCN-Attention. The experiments were executed on URFD Da taset, Le2i Dataset, and a combined, URFD-Le2i dataset to analyze and evalu ate their performance. According to the experimental results, the method that combines GMFlow-CNN-TCN-Attention achieved better performance than the other proposed models. This model obtained test accuracies of 100% on the URFD dataset, 92% on the Le2i dataset, and 94% on URFD-Le2i dataset. These results point out that the presented method is capable of effectively cap turing both spatial features and temporal features required for fall detection. This approach provides useful insights for developing effective real-time vi sion-based fall detection applications.
Professor, National Institute of Technology Raipur, India.
Narendra D. Londhe is presently working as Associate Professor in the Department of Electrical Engineering of National Institute of Technology Raipur, Chhattisgarh INDIA. He completed his B.E. from Amravati University in 2000 followed by M.Tech and Ph.D. from Indian Institute of Technology... Read More →
Authors - Kalva Yamini, Kapilesh C, Hari Kishore R, Giri Karthick S Abstract - Phishing attacks remain among the most prevalent cybersecurity threats, exploiting deceptive URLs that imitate legitimate domains. Traditional blacklist and heuristic-based methods fail to detect zero-day phishing URLs, leaving users exposed to novel attack vectors. This paper presents 3P-VAD (Three-Phase Verification and Detection), an AI-powered system for real-time URL classification integrating three complementary layers: (i) threat intelligence dataset lookup against live feeds, (ii) multi-engine verification via the VirusTotal API aggregating results from 70+ security vendors, and (iii) a Convolutional Neural Network (CNN)-based zero-day detection model operating exclusively on URL character sequences. A selection-based scanning mechanism enables on-demand URL verification, enhancing user privacy by preventing inadvertent submission of sensitive internal URLs to third-party services. Evaluated on 2 million URLs, the framework achieved 95.0% accuracy, 94.5% precision, 86.0% recall, and 90.0% F1-score on the CNN zero-day component, with 100% combined detection rate across all three phases. Ablation experiments confirm non-redundant, complementary coverage.
Authors - Farai C. Jonha, Arthur Ndlovu, Mainford Mutandavari Abstract - This study presents a systematic review on the use of deep learning and density-based techniques for explainable segmentation of banking customers. We analyze 71 peer-reviewed papers published between 2015 and 2025 to investigate their methodological trends, validation approaches, and the degree of incorporation of interpretability into proposed models. Our findings suggest that autoencoders and variational autoencoders provide better separation of clusters than models using raw data. In terms of clustering methods, density-based clustering algorithms perform better than clustering algorithms based on centroids since banking data exhibit highly skewed and non-Gaussian patterns. We also observe a common deficiency in explainability, with less than 26% of the re-viewed papers considering approaches such as SHAP or LIME. Furthermore, considerations of external validity, operational governance, regulation, and scalability of implementation are rare. We therefore propose an explainable customer segmentation (XCS) framework based on deep representation learning, density-based clustering, post-hoc explainability, and an operationally ready pipeline that is suitable for use in regulated banking environments
Authors - Mohammed Sulaiman I, Shreevatsa DS, Kavitha Sooda, Revanth L, Dhanush M Abstract - The unintentional release of API keys, tokens, and any other credentials in the source code is an obvious security threat to contemporary software development. Old rule-based scanners produce too many false positives and cannot scan through obfuscated secrets or secrets that are unknown. This paper introduces AV-SHIELD (Automated Vulnerability Scanning Hybrid, Implementing integrated Leakage Detection) which is a hybrid framework that brings together pattern matching and machine learning to identify credential leaks in real time. The system serves to monitor development spaces in event-driven fashion and scan repositories in GitHub up to size limitations. One uses a Random Forest type of classifier, which is trained on entropy based features to combatSecret vs Benign strings and a risk scoring engine which gives priority to create alerts. Records of the identified exposures are archived in a fingerprint-tracked vault, batch-processed into mail notifications, and include professionally-formatted PDF records. A trade analysis using an interactive Streamlit dashboard allows viewing trends of exposure, provider profiles, and risk allocations. The synthetic data generated has demonstrated a high precision and recall rate that is much lower than the explanation of the uses of regex alone, tested through experimental evaluation. The framework was implemented as a systemd service, which shows its applicability to the enterprise DevSecOps pipelines.
Authors - Alfred ADINSI, Pelagie HOUNGUE Abstract - This systematic review evaluates AI-based techniques for rice dis-ease detection with a focus that existing surveys have not adopted: their deploy-ability in West African smallholder conditions, using Benin as the reference case. Based on 220 studies selected from 390 Scopus publications (2019–2025) via PRISMA, it goes beyond performance benchmarking to assess what actually works under resource constraints. Rice blast (70.9% of studies), brown spot (60.9%), and bacterial blight (44.5%) dominate the literature. Deep Learning accounts for 64.5% of approaches, hybrid methods for 21.8%, and classical Machine Learning for 13.6%. Mean accuracy reaches 94.2% for pure Deep Learning and 95.8% for hybrid architectures. Res-Net+ViT (96.4% ± 2.1%) and CNN+SVM (94.1% ± 4.1%) are the strongest per-formers, but performance alone is not the right metric for Benin. While 85% of studies apply to tropical climates, only 30.5% propose solutions running on limited hardware. Three approaches clear both bars: MobileNet+SVM (89.4%), optimized YOLOv8 (89.2%), and ResNet-based Transfer Learning (91–94% after fine-tuning). That AI can detect rice diseases accurately is no longer in question. The harder problem is which systems beninese farmers and extension agents can actually use. This review provides an answer.
Authors - I Gusti Ayu Agung Dewi Sucitawathi Pinatih, Jonathan Jacob Paul Latupeirissa Abstract - The objective of this study is to examine and analyze the integration of technology, governance, and sustainability in the context of e-government and public services, with a particular focus on the implementation of these three dimensions at the global and local levels, specifically in the Province of Bali. This study employs a Systematic Literature Review (SLR), beginning with the identification of relevant keywords such as “e-government,” “public service,” and “sustainabil-ity,” which were validated using WordCloud. Next, strict inclusion and exclusion criteria were used to select articles. These criteria included relevance to the topic, year of publication (2016-2026), and the journal’s peer-review status. Initial identification, screening of titles and abstracts, and in-depth reading of articles were part of the article selection process. The research findings indicate that in the digital transformation of the public sector, technology, governance, and sustainability are interrelated, and Bali serves as an example of how the integration of these three dimensions is reinforced by local values such as Tri Hita Karana and the subak system. These findings underscore that the digitization of public services in Bali will succeed if the principle of sustainability is applied in tandem with technology, governance, and local culture.
Authors - Aymane Chekira, Aziz Hmioui Abstract - The rapid expansion of digital technology in recent years has significantly changed the way international supply chains (SCs) are structured, operated, and how well they perform. Among these transformations, blockchain has grown to be a major enabler for addressing continuous concerns with transparency, traceability, collaboration, and trust throughout supply chain networks. As companies seek more and more to raise supply chain performance and sustainability, scholarly investigations of blockchain-based supply chain management have grown dramatically. Descriptive and content analysis of co-occurrence key-words using Biblioshiny and VOSviewer software revealed the main research subjects and their linkages across 145 peer-reviewed Scopus-indexed publications spanning 2019–2026. Scientometrically speaking, this study examines this expanding body of research. The results point to two primary research directions: (i) how blockchain uptake influences organizational performance and supply chains, and (ii) how transparency, traceability, decision-making, and sustainable development enabled by blockchain are present in supply chains. The data analysis reveals that blockchain technology is a key and unifying feature that connects performance improvement with the goals of governance and sustainability. It emphasizes new ways for more investigation in blockchain-enabled supply chain performance and offers a systematic overview of the intellectual environment of blockchain research in supply chain management, as well as comprehensible in-sights on its thematic evolution.
Authors - Putu Putri Prawitasari, Shefali Saluja, Jonathan Jacob Paul Latupeirissa Abstract - Financial statements are vital for conveying a company's performance and financial health, yet fraudulent financial reporting remains a significant concern, especially involving fraud hexagon schemes. This study investigates the integration of advanced technologies to combat fraud hexagon schemes and improve auditing effectiveness in the digital era. Through a comprehensive literature review of academic sources from the Scopus database, this research identifies the limitations of traditional auditing in detecting complex fraud patterns. Findings reveal that the adoption of technology-based tools such as data analytics, artificial intelligence, machine learning, and blockchain enhances auditors’ ability to detect anomalies and suspicious activities more efficiently and accurately. Furthermore, combining these technologies with robust corporate governance and auditor expertise strengthens fraud prevention mechanisms. The study concludes that leveraging digital innovations within a holistic fraud detection framework significantly advances audit quality and fraud mitigation strategies in contemporary financial environments.
Authors - Valencia Vannessa Taslim, Melissa Anastasia, Shalva Andena Rizaldi, Tiurida Lily Anita Abstract - Online hospitality reviews provide valuable insights into guest experiences, service quality, and operational performance. However, the unstructured and noisy nature of review text makes large-scale analysis difficult, especially for Indonesian-language reviews that often contain informal expressions, abbreviations, spelling variations, and inconsistent sentence structures. Although sentiment analysis has been widely applied in hospitality research, studies focusing on Indonesian-language hospitality reviews remain limited, and few have presented a reproducible Natural Language Processing (NLP) workflow for multiclass sentiment classification. This study proposes a reproducible Indonesian NLP pipeline for classifying hospitality reviews into positive, neutral, and negative sentiment categories. The workflow integrates review collection, sentiment annotation, Indonesian text preprocessing, TF-IDF feature extraction, and super-vised classification using Naïve Bayes, Support Vector Machine (SVM), and Logistic Regression. The dataset consists of 450 Indonesian-language hotel reviews collected from Google Reviews across three hotel segments: budget, mid-scale, and upscale. The experimental results show that SVM achieved the best overall performance, with 91.78% accuracy, 91.35% precision, 91.78% recall, and 91.50% F1-score, outperforming Naïve Bayes and Logistic Regression under the same experimental setting. These findings indicate that classical machine learning, when supported by systematic preprocessing and consistent feature representation, remains highly effective for Indonesian hospitality review analytics. This study contributes a practical and reproducible baseline for Indonesian-language sentiment classification and provides a foundation for future intelligent review monitoring systems in the hospitality sector.
Authors - Jolou Vincent M. Jala, Everly A. Nacalaban, Nenon Roy A. Sandinao, Erlinda D. Rivera, Hilfiger L. Cubarrubia Abstract - This article explored the adoption of Artificial Intelligence in Financial Management Systems of Higher Education Institutions (HEIs) by utilizing a systematic review of related literature. The study focuses on reviewing pre-sent literature on Artificial Intelligence adoption in financial management systems, recognizing the benefits of AI integration, scrutinizing the challenges and barriers to implementation, and offer recommendations for effective and successful AI integration in HEIs. The findings disclosed that artificial intelligence has the capability to meaningfully enhance financial management systems in Higher Education Institutions through automated financial reporting systems, budgeting forecasting and predictive analytics, fraud detection and risk management, and expense tracking and optimization. Adoption of Artificial Intelligence improves efficiency, enhances accuracy, provides better decision-making and cost optimization. More-over, it enhances operational efficiency by systematizing monotonous financial tasks, enhances accuracy by plummeting human faults, supports better decision-making through actual financial data and predictive analytics, and helps to long-term cost optimization and financial sustainability. These improvements permit institutions to alter from manual and volatile financial management routines toward more data-driven, calculated and strategic, financial planning and re-source provision. Conversely, the study also found several challenges that deter AI adoption in Higher Education Institutions, specifically in developing countries such as the Philippines. These challenges include high initial investment and maintenance costs, limited technical skills among staff, data privacy and cybersecurity risks and resistance to organizational change. Numerous HEIs are still in the developing stage of digital transformation and depend chiefly on enterprise systems and basic accounting rather than advanced Artificial Intelligence technologies. The article concludes that successful AI in- corporation requires institutional readiness, strategic planning, capability building, infrastructure progress, and robust data governance policies to completely maximize the advantages of Artificial Intelligence in financial management systems.
Authors - Denver Novencido Abstract - An organization’s operational efficiency, productivity, and reliability can be adversely affected by using manual-based systems. Some of the issues associated with using a manual-based approach include inefficient processes, inconsistent documentation, difficulty in monitoring and validating records, and limited accessibility. The development of information systems provides a solution to address the limitations and challenges of a manual-based approach in organizations. This study presents the design and implementation of a cloud-based information system integrated with decision support capabilities to streamline organizational operations, enhance data storage and retrieval, and facilitate strategic planning. The system was created using the Agile Unified Process (AUP) software development methodology. Evaluation results indicate strong compliance with ISO software quality standards, making it a suitable tool for managing organizational operations.
Authors - Felix Kabwe, Jackson Phiri Abstract - This study explores how blockchain-based Identity and Access Management (IAM) systems can enhance the security and efficiency of Digital Financial Services (DFS). As DFS environments grow more complex and involve multiple stakeholders, traditional IAM systems face challenges such as centralization, limited interoperability, and scalability constraints. Blockchain offers a compelling alternative by enabling decentralized, transparent, and tamper-resistant identity management. The study compares three main IAM models: centralized systems supported by blockchain, federated identity management, and Self-Sovereign Identity (SSI). Using the Technology-Organization-Environment (TOE) framework alongside a semi-quantitative scoring approach, the research evaluates these models across key factors including security, privacy, usability, scalability, governance, cost, and regulatory alignment. The findings highlight clear trade-offs. Centralized systems excel in performance, cost efficiency, and regulatory compliance but are vulnerable to single points of failure. Federated models strike a balance by improving interoperability and user experience, though they introduce governance complexity. SSI provides strong privacy and user control but faces challenges in usability, scalability, and regulatory acceptance. Overall, no single model fully meets DFS needs. Federated systems are currently the most practical, while hybrid federated–SSI approaches offer the most flexible, scalable, and user-focused solution.
Authors - Md. Monowar Hossain, Fahima Hossain, Md. Shahidul Islam, Md. Tanvir Ahmed, Reduan Ahmed Abstract - This automated image captioning is on one hand a Computer Vision (CV) and Natural Language Processing (NLP) application, but on the other hand, conventional CNN-RNN models suffer from feature loss and long-range dependency. The proposed model in this study is a parameter balanced multi-modal model that consists of a dual-encoder network which combines Effi-cientNet-B4 for hierarchical features and MobileNetV2 for geometric efficiency, as well as a multi-head Transformer decoder. The model was evaluated on Flickr8k, and tested with a dynamic scalar weight mechanism and teacher-forced optimization, the BLEU-1 was 0.5774 and METEOR was 0.4129. Interestingly, the ablation results also showed that although the dual-encoder method is competitive, the pathway of the standalone MobileNetV2 is slightly better than the fused pathway in terms of BLEU-4 (0.2284 vs. 0.20). This indicates that the pathway may be redundant during the concatenation process. This study validates the possibility of using Transformer decoders instead of RNN bottle-necks and offers important considerations for the optimization of real-time feature fusion for vision tasks.
Authors - Aarthi R, Aniketha Prasad, Dhamini Manoj, Manasvi G, Meghaa Sunil Abstract - Early and accurate diagnosis of dermatological disorders remains one of the main issues in clinical dermatology, especially with regard to diseases with similar appearances. Despite the achievements of deep learning methodologies in the classification of cutaneous lesions with the help of images, structured clinical metadata is not used to the fullest, despite its significant diagnostic potential. In a practical clinical setting, dermatologists do not solely use visual evaluation of the case but also use patient-specific metadata, which includes age, lesion progression, pruritus, hemorrhage, anatomic location, prior biopsy, and family history. The current study presents a fully explainable, metadata based, multi-class classification of skin diseases, using the PAD-UFES-20 database, and concentrated on 6 distinct diagnostic categories. Although the dataset is dermoscopic, the predictive quality of formal metadata variables are mainly under consideration in the present work. The explainability analyses revealed that biopsy status, elevation, itch, region and age are attributes that have significant effects on the classification results. However, empirical evidence shows that the reduced model consisting of the premier five features lowers accuracy, which highlights the importance of a thorough combination of metadata features to determine skin disease rather than limited combination. Comparative studies indicate that the Multi-Layer Perceptron shows an improvement in a model performance with a corresponding increase of the number of selected features. The suggested framework thus highlights interpretability in line with predictive efficacy thus enhancing the importance of transparent artificial intelligence systems in medical decision-making.
Authors - Veeravalli Sri Satya, Anjan Babu G Abstract - Human-computer interaction with smart consumer electronics predominantly requires physical peripherals, which introduce limitations regarding hardware degradation, shared-surface hygiene, and usability in hands-free environments. Voice-activated systems provide an alternative but exhibit high latency and degraded performance under ambient noise. This paper presents a multi-layered touchless gesture control framework that translates human hand kinematics into direct system actuation. The architecture utilizes a standard web camera and the Google MediaPipe framework to extract 21 three-dimensional hand landmarks in real time. To bypass the computational bottlenecks of traditional Convolutional Neural Networks (CNNs), the system employs a custom heuristic algorithm to classify eight distinct static and dynamic gestures by analyzing the geometric relationships between finger joints. The framework processes these classifications locally and actuates Android-based Smart TVs over Wi-Fi utilizing Android Debug Bridge (ADB) protocols [11]. Evaluated in a controlled environment, the pipeline achieved an average processing time of 35 milliseconds per frame (approximately 30 frames per second) with a network transmission delay of 50 to 80 milliseconds. The results suggest that computationally lightweight computer vision models, when paired with structured state-machine logic, can effectively replace physical remote controls without requiring dedicated GPU hardware.
Authors - Sachin Ramling Jadhav, Rajveer Nandkar, Srushti Rajput, Rajvardhan Desai, Gunjan Ramteke,Samruddhi Rajput Abstract - Dealing with city problems like cracked roads, trash piles, leaks in pipes, or dark lamp posts keeps urban teams busy. When fixes depend on old paper methods, pieces of info get lost, trust dips, responses drag. A new setup steps in - CCIRS - running through a basic website made with PHP tools. Instead of guessing what comes first, supervisors follow a clear score called PI, shaped by how bad things look, where many reports cluster, plus how long issues wait. Behind the scenes, staff watch live updates, study trends, trace progress using their control view online. Half a year of testing in three city areas of Pune cut response times by 59.0%. Because of this change, meeting service targets got better by nearly half. Old ways of handling issues were clearly outperformed. Math behind sorting locations was built and tested. Ranking urgency used formulas that matched real outcomes well.
Authors - Mrityunjaya Chavannavar, Melita Simoes , Nikhil Shetty , Chirivella Vishal Abstract - Over the years, there is a rapid growth of social media-based financial content. Finfluencers have been emerging as influential sources that provide investment information to young retail investors. This research is inclined towards understanding the influence of finfluencers on numerous behavioural biases that include herd mentality, overconfidence, and FOMO. This study also examines their influence on decision-making when it comes to investments and the overall risk perception in the current digitally enabled investment landscape. There is interplay between social media platforms, financial influencers, and behavioural biases and can be observed among young retail investors in India. Most traditional theories in finance assume that a majority of investors behave rationally while behavioural finance acknowledges the impact of cognitive and emotional biases influence investment decisions. This quantitative study makes use of a descriptive-analytical approach. The primary data used here was gathered with the help of structured online questionnaires distributed to 120 young retail investors. Data analysis was carried out with the help of IBM SPSS Statistics. Tests such as correlation analysis, multiple regression models, and ANOVA with post-hoc Tukey HSD were undertaken. Findings showed that general social media usage frequency had no significant relationship with the four behavioural biases examined. Perceived credibility of finfluencer content demonstrated significant negative relationships with all four biases (overconfidence: β = -0.387, p = 0.001; herding: β = -0.252, p = 0.044; confirmation: β = -0.321, p = 0.006; availability: β = -0.354, p = 0.003). This indicates that high-quality financial influencers may serve a corrective rather than amplifying function. Indiscriminate following of numerous finfluencers positively predicted confirmation bias (β = 0.191, p = 0.025). Investors with over five years of experience revealed significantly lower biases. This study can be used for better investor protection, and financial literacy initiatives and can be embedded in various regulatory frameworks.
Authors - April L. Macasieb-Gumnad, Roberto M. Arguelles Abstract - The study focuses on transformational leadership, entrepreneurship, and sustainability in higher education. Using Saint Louis University (Philippines) as a case study, the purpose was to (1) identify the role transformational leadership has in developing (or affecting) the characteristics of an entrepreneurial university, (2) identify how transformational leadership fosters sustainable innovation, and (3) assess the effect entrepreneurial university characteristics have on achieving sustainable outcomes. This quantitative research used three different instruments that were previously validated (HEInnovate Questionnaire; Sustainability Assessment Questionnaire; and Survey of Transformational Leadership) to gather data from a sample of 795 respondents at SLU and analyzed the resulting data using Spearman-rank correlation analysis and simple linear regression. This study provided practical applications to the literature on higher education management through empirical evidence of relationships between types of leadership styles, achievement of SDGs, organizational structures/models/characteristics, and sustainability of innovation in higher educations.The SLU CARES Innovation Framework was proposed to provide actionable insights for academic and administrative leaders seeking to align Catholic educational missions with contemporary demands for innovation and sustainability.
Authors - Darrel A. Cardana, Ethel Zean M. Anosa, Angeline B. Elegio, Jes Maries Mendez, Ivy Corazon Mangaya-ay Abstract - Agri-Aqua Technology Business Incubators (ATBIs) play an important role in promoting innovation, entrepreneurship, technology commercialization, and institutional collaboration within higher education institutions. This study assessed the operational performance and institutional development of the BISU Agri-Aqua Technology Business Incubator (ATBI). Specifically, the study evaluated the accomplishments of the incubator in terms of personnel capacitation, partnership and linkage development, awareness and promotional activities, incubation services, technology incubation initiatives, intellectual property generation, and policy institutionalization. The study also examined the capacitybuilding activities, partnership initiatives, intellectual property outputs, and the problems and strategic solutions encountered during implementation. The study employed a descriptive-evaluative research design utilizing documentary analysis of the official accomplishment report and supporting institutional documents of the BISU ATBI. Frequency counts, percentage analysis, and thematic analysis were utilized in analyzing the collected data. The findings revealed that the BISU ATBI successfully implemented several operational and institutional initiatives. The incubator conducted seventeen (17) trainings and workshops, forged twelve (12) MOUs with incubatees and six (6) institutional partnerships, conducted eight (8) awareness seminars, developed ten (10) business plans, filed ten (10) trademarks and five (5) copyrights, and enrolled seventeen (17) incubatees in the incubation program. However, only two (2) technologies were successfully co-incubated despite the target of ten technologies, indicating challenges in technology commercialization and adoption. The study also identified regulatory hurdles, technology readiness concerns, partnership issues, and low technology adoption as major implementation challenges. Overall, the findings indicate that the BISU ATBI established a strong operational and institutional foundation for technology business incubation, although continuous enhancement of commercialization and technology adoption initiatives remains necessary.
Authors - Nguyen Quoc Cuong, Nguyen Ha, Mai Thi Bich Ngọc Abstract - The proliferation of short-form video platforms has reshaped consumer decision-making, yet how electronic word of mouth (eWOM) attributes influence Generation Z fashion shopping intention in emerging markets remains underexplored. Grounded in the Information Adoption Model (IAM) and Attitude–Intention framework, this study examines the impact of TikTok-based eWOM on fashion shopping intention among Vietnamese Generation Z consumers. Using PLS-SEM analysis of 263 valid survey responses, results reveal that eWOM Information Quality and Credibility significantly predict Attitude toward eWOM, while Information Usefulness is the strongest predictor of eWOM Adoption; Information Quantity exerts a positive but weaker effect. Both Attitude and Adoption significantly influence Fashion Shopping Intention, with Attitude as the dominant predictor, and mediation analysis confirms their roles as key intervening mechanisms. These findings extend the IAM to short-form video and social commerce contexts, demonstrating that Generation Z engages in evaluative, quality-oriented content processing rather than responding passively to volume. Practically, results offer actionable guidance for fashion marketers to prioritize authentic, credible, and informative eWOM strategies on TikTok.
Authors - Rakhmonova Nargiza Rashidovna, Rajapov Shukhrat Zaripbaevich Abstract - The growing volume of international trade is increasing pressure on road border customs posts, making their operational efficiency a key factor in facilitating foreign trade. Chronic congestion, long vehicle queues, and procedural delays at land border crossings hinder logistics efficiency and increase trade costs. Digitalization is increasingly viewed as a strategic solution for modernizing customs administration while ensuring effective control and economic security. This study examines ways to further improve the efficiency of road border customs posts through digitalization, using the case of Uzbekistan. The analysis is based on data from 322 road border customs posts and employs economic and statistical methods, including regression analysis and structural equation modeling (SEM). The model assesses the impact of human resources, infrastructure capacity, and digital inspection technologies—specifically, the number of employees, traffic lanes, inspection and verification complexes (ISC and Z-portal), passenger flows, and reported violations—on daily vehicle traffic volumes. The results consistently show that human resources are the most significant factor in customs post efficiency. An increase in the number of employees has a strong and statistically significant positive effect on daily vehicle flow across all parameters of the model. In contrast, the expansion of physical infrastructure, measured by the number of traffic lanes, shows a negative or weakly significant relationship, indicating that infrastructure alone does not guarantee increased throughput. Digital control systems show a positive but statistically insignificant effect, suggesting incomplete integration into operational processes. The results indicate that to achieve significant efficiency gains, digitalization must be combined with effective human resource management and organizational optimization. Policy measures should prioritize capacity building, intelligent traffic management, and deeper integration of digital systems to reduce congestion, speed up logistics, and improve conditions for foreign trade.
Authors - Al John A. Villareal, Jaime M. Samaniego Abstract - Potholes significantly impact road safety, vehicle performance, and infrastructure maintenance, particularly in developing countries where monitoring systems remain largely manual. This study presents the design and implementation of an Android-based mobile application that utilizes sensor fusion and deep learning for real-time pothole detection and severity analysis. The system integrates a YOLO-based object detection model with smartphone sensors, including accelerometer, gyroscope, and Global Positioning System (GPS), enabling simultaneous visual and motion-based detection. A dataset consisting of 9,253 road surface images containing 16,123 pothole annotations was used for training and evaluation using a 70:20:10 dataset split for training, validation, and testing. Among the evaluated models, YOLO11s achieved the highest mAP@50– 95 value of 54.2%. However, YOLO26n was selected and implemented in the developed Android application due to its competitive detection performance, compact 5.2 MB model size, and suitability for real-time mobile deployment. Field testing across four road segments covering 18.97 kilometers resulted in 130 detections, of which 84 were verified potholes and 46 were false detections, yielding a verification rate of 64.62% and a false detection rate of 35.38%. The system recorded an average detection density of 6.85 potholes per kilometer. Results demonstrate that integrating deep learning and sensor fusion in a mobile platform provides a scalable and cost-effective solution for automated road condition monitoring and intelligent transportation systems.
Authors - Nu Yin Khaing, Win Lelt Lelt Phyu Abstract - Water quality monitoring is essential for sustainable aquaculture management and fish health assessment. However, monitoring a large number of physicochemical parameters in creases sensor deployment costs and system complexity. While traditional machine learning ap proaches struggle with complex, nonlinear relationships among water quality variables, feature reduction can optimize system efficiency. This study proposes a hybrid machine learning and deep learning framework to achieve accurate, cost-effective water quality classification using the Water Quality Dataset (WQD). The framework integrates Random Forest (RF) and XGBoost for feature selection with Support Vector Machine (SVM) and Long Short-Term Memory (LSTM) models as classifiers, evaluating four hybrid combinations (XGBoost+SVM, XGBoost+LSTM, RF+SVM, and RF+LSTM) across subsets of 10, 7, and 5 features. Experimental results demon strate that hybrid deep learning architectures consistently outperform traditional machine learn ing methods. Specifically, XGBoost+LSTM and RF+LSTM achieved the highest classification accuracy of 96.05% using 10 selected features, while maintaining reliable performance at lower dimensions. Furthermore, Shapley Additive explanations (SHAP) analysis was applied to en hance model interpretability, identifying Dissolved Oxygen (DO), Turbidity, BOD, H2S, and Nitrite as the most important attributes. Ultimately, the proposed framework minimizes sensor requirements and provides an accurate, interpretable, and economically viable solution for aqua culture water quality monitoring systems.
<b>Authors - </b>Mahi Shah, Sachin Pande, Sumitra Jakhete, Emmanuel Mark<br /> <b>Abstract - </b>Brain-Computer Interfaces (BCIs) operate as systems that translate brain signals into digital commands. They provide a non-muscular channel of communication for individuals with profound motor disabili ties. Cerebral Palsy (CP) is a neurological condition that impairs move ment and muscle tone, frequently making physical or verbal expression difficult. This paper reviews the current state of BCI technology and, building upon these insights, introduces a framework for a non-invasive, low-cost BCI communication system tailored specifically for children with CP, addressing the limited accessibility of assistive communication technologies in low-resource environments. The proposed seven-stage framework targets these ongoing challenges by incorporating OpenBCI hardware, adaptive signal processing, and gamified interfaces. This processing pipeline converts neural signals into structured communication outputs, enhancing accessibility and engagement for CP children. To assess the feasibility of the proposed framework, an offline analysis was conducted using a publicly available EEG dataset. A Linear Discriminant Analysis (LDA) classifier y a classification accu racy of 62.5% and an Information Transfer Rate (ITR) of 11.4 bits/min, demonstrating the computational viability of the approach. The modular design offers scalability, though its efficacy requires further validation in real-world pediatric settings. In summary, this work bridges theoretical insights with practical innovation, offering a promising step toward empowering CP children. While limitations in real-world testing remain, the framework lays a foundation for future refinements. Successful implementation could significantly improve independence and quality of life, marking a milestone in inclusive assistive technology.
Authors - Apolinar P. Datu, Barnard J. Maraon, Rommel H. Orquiza, Cristopher T. Takano, Olivia L. Yosa, Mark Joseph G. Cruz, Jeferson D. Talisayon Abstract - This study examines the integration of multilingual customer service skills into workforce development programs for hospitality management students. In today’s globalized environment, the hospitality industry serves guests from diverse cultural and linguistic backgrounds, making effective communication an essential skill. Using a quantitative approach, the study gathered data from 100 hospitality students to assess their communication skills, performance, and confidence in multilingual settings. The findings reveal that most students already possess basic multilingual abilities, particularly in English and Filipino, which serve as a strong foundation for further development. Results also show that the current curriculum includes elements of multilingual training, contributing to students’ overall competence. However, while students demonstrate satisfactory performance and confidence, there is still a need for increased practical exposure and strengthened training programs. Furthermore, the study highlights that students generally meet industry expectations but may benefit from more structured and continuous training to enhance their real-world readiness. Overall, integrating multilingual customer service skills significantly supports the development of hospitality students, preparing them for the demands of a culturally diverse workforce and improving their competitiveness in the global hospitality industry.
Authors - Sanjana Priyadarsini, Choudhary Aman Kumar Roy, Ashlesha Shree Bajpai, Rajdeep Banerjee, Shivali Sharma, Ranjita Kumari Dash Abstract - Today, machine learning methods are quickly being adopted in healthcare. In numerous instances, it has been observed that datadriven approaches have increased reliance of medical data analysis and disease detection by about 60-70%. It is important to diagnose cardiac arrhythmias early using electrocardiogram (ECG) analysis, as timely diagnosis can prevent severe complications and loss of life. Most ECG datasets are however not balanced, with normal beats by far outnumbering abnormal ones and causing the models to underperform on rare but significant cases. In this work, Logistic Regression is used as a baseline model. To correct this imbalance, Class weighting and Synthetic Minority over-sampling Technique (SMOTE) are applied. These techniques help the model detect rare heartbeat patterns more reliably and miss fewer abnormalities. This paper shows that addressing class imbalance can make ECG-based classification systems more accurate and clinically valid.
Authors - Windy Permata Suyono, Dwi Handarini, Eka Septariana Puspa, Surya Anugrah, Nuramalia Hasanah, Ratna Anggraini, Sabo Hermawan, Rio Firnanda Abstract - This study aims to develop an AI-Based Digital Audit Maturity Frame work integrated with SPIP to support Smart SME Governance and Sustainable Development Goal 9 (SDG 9). The study employs a systematic literature review approach by analyzing 50 relevant articles published between 2020 and 2026 re lated to artificial intelligence in auditing, digital audit maturity, internal control systems, smart governance, SMEs, and sustainable innovation. The findings in dicate that artificial intelligence technologies significantly improve audit effec tiveness, governance transparency, operational efficiency, and organizational re silience. The study also reveals that digital audit maturity and SPIP-based internal control systems play important roles in supporting sustainable digital transfor mation and adaptive governance within SMEs. Based on the literature synthesis, this study proposes a conceptual framework consisting of input factors, digital transformation processes, digital audit maturity levels, SPIP-based internal con trol systems, smart SME governance, and SDG 9 achievement. The proposed framework contributes theoretically by integrating technological capability, gov ernance systems, internal control mechanisms, and sustainability perspectives into a unified governance model. Practically, the framework provides guidance for SMEs, policymakers, auditors, and digital transformation practitioners in strengthening sustainable AI-driven governance implementation.
Authors - Milind Nemade, Khush Chheda, Rahul Dhanak, Durgeshkumar Dubey Abstract - The problem of meeting productivity continues to prevail in the current era in multilingual environments due to frequent language switching between speakers. Most of the existing frameworks for meeting intelligence primarily focus on automatic transcription and lack significant support for Indic languages, speaker identification, and task extraction. Additionally, many of these frameworks depend on metadata associated with specific platforms and, therefore, cannot be used in any offline environment or even on other platforms. In this work, we present a scalable and platform-agnostic framework for meeting intelligence which can automatically an alyze meetings post factors by leveraging speech recognition, speaker identification, and contex tual analysis. The system leverages multilingual and code-switched transcription capabilities of Sarvam AI, generates speaker embeddings using ECAPA-TDNN, and then uses Large Language Models for context-based analysis. Two different strategies for speaker identification are dis cussed in this paper such that they do not need any platform-based metadata while improving the attribution accuracy. We have further developed an asynchronous framework for extracting tasks, assigning tasks, and notifying about them. Experimental results indicate enhanced transcription accuracy as well as speaker identification accuracy in Hindi-English code-switching cases. Fu ture work will focus on implementing advanced privacy protection and end-to-end encryption mechanisms for secure storage and processing of meeting recordings and metadata.
