Journal of Information Systems and Informatics

ERP Adoption in Higher Education: A TAM-Based Analysis of Botswana’s Technical University

Boitshoko Effort Otlhomile, Ofaletse Mphale, Joyce Mosinki — Fri, 05 Dec 2025 00:00:00 +0700

This study investigates ERP adoption at a technical university in Botswana using the Technology Acceptance Model (TAM). It examines how Perceived Usefulness (PU), Perceived Ease of Use (PEOU), and Behavioural Intention (BI) influence Actual System Use (AU). Data were collected from administrative staff using a structured survey and analyzed using regression analysis. The results show that PEOU significantly influences both BI (R² = 0.964, p = 0.0029) and PU (R² = 0.864, p = 0.022), indicating that system usability is crucial for ERP adoption. Furthermore, PEOU positively impacts PU (R² = 0.817, p = 0.035), and BI strongly predicts AU (R² = 0.821, p = 0.034). These findings highlight the importance of user-friendly interfaces, comprehensive training programs, and institutional support to ensure successful ERP implementation. The research provides valuable insights for universities aiming to enhance operational efficiency, streamline data management, and improve decision-making processes through effective ERP adoption, particularly in developing countries like Botswana.

Enhancing Mobile Library App User Experience Using HCD and Usability Metrics

Fri, 05 Dec 2025 00:00:00 +0700

This study analyzes the improvement of usability metrics and the correlation between usability variables and user experience in the development of a mobile-based digital library application. Using the Human-Centered Design (HCD) approach, the study employed Concurrent Think Aloud (CTA) and Post-Study System Usability Questionnaire (PSSUQ) for usability evaluation. The focus was on four main usability metrics: effectiveness, efficiency, satisfaction, and learnability. The study involved 100 respondents from IT Del students. The results showed significant improvements in all usability metrics: effectiveness increased from 88% to 100%, efficiency rose from 0.087 to 0.148 goals/second, satisfaction improved from 82.05% to 87.05%, and learnability improved with the number of failed tasks reducing from four to zero. Multiple linear regression analysis revealed a strong positive correlation between usability metrics and user experience, with an R² value of 0.665, meaning 66.5% of the variation in user experience can be explained by the usability metrics. All usability metrics positively contributed to improving user experience. These findings confirm that applying HCD and systematic usability evaluation can significantly enhance the quality of digital applications, particularly for mobile-based libraries, and offer valuable insights for the design of digital library apps in higher education contexts.

A Unified Framework for Theoretical and Experimental Evaluation of Classical and Modern Sorting Algorithms in Real-Time Systems

Japheth Kodua Wiredu, Stephen Akobre, Iven Aabaah, Umar Adam Wumpini — Tue, 09 Dec 2025 14:31:08 +0700

This paper presents a theoretical and experimental evaluation of eight popular sorting algorithms HeapSort, QuickSort, MergeSort, Parallel MergeSort, TimSort, IntroSort, Bitonic Sort, and MSD Radix Sort—assessing their suitability for real-time computing environments. The study combines algorithmic analysis with large-scale benchmarks across various input distributions (random, almost sorted, reverse-sorted) and data scales, focusing on execution time and memory usage. Results show that hybrid and adaptive algorithms outperform classical ones. TimSort had the shortest execution times (as low as 1.0 ms on sorted data), and IntroSort showed consistent performance across data types (11-13 ms on random inputs) with minimal memory (<7.90 MB). HeapSort maintained predictable O (n log n) behavior, suitable for hard real-time constraints, while QuickSort and MergeSort had lower latency but higher memory usage. These findings are significant for latency-sensitive applications like high-frequency trading and sensor data processing. The study recommends using hybrid algorithms like TimSort and IntroSort for general-purpose workloads, providing evidence-based guidance for real-time system design.

Perceptions and Key Factors Influencing the Concept of Smart Bangladesh

Md. Monowar Uddin Talukdar, Khayrul Alam, Shakila Ferdous, Ripa Akter — Tue, 09 Dec 2025 15:39:05 +0700

This study explores the concept of a smart Bangladesh, focusing on the roles of smart citizens, government, society, and economy. It argues that leveraging technology is not limited to digitizing government services but also involves transforming the interactions between citizens, society, the economy, and the government. The findings highlight the significant influence of smart citizens and a smart economy on this concept, emphasizing its relevance for both developing and underdeveloped countries. A total of 179 responses were collected using random sampling, ensuring comprehensive coverage for structured interviews based on the Likert scale. The impact was analyzed using inferential statistics with SmartPLS (Version: 4.0.9.9). Bibliographic data spanning from 2018 to 2023 were visualized using VOSViewer, mapping 214 pieces of literature from the Web of Science (WoS) database to support the concept. Structural Equation Modeling (SEM) and Exploratory Factor Analysis (EFA) yielded an R-square value of 51%. The results confirm the acceptance of hypotheses H1 (β = -0.057, t = 0.730, p > 0.233) and H4 (β = 0.603, t = 5.459, p < 0.000), showing a direct effect on the smart concept. This study presents a holistic approach to sustainable development through technological transformation, consolidating research across smart domains like healthcare, education, agriculture, payments, and grids.

AI-SEC-EDU Conceptual Framework: Securing E-Learning in Low-Income Countries’ Higher Education Institutions

Justine Mukalere, David Andrew Omona, Agatha Flavia Ikwap — Tue, 09 Dec 2025 16:19:59 +0700

The evolving digital threat landscape, characterized by sophisticated AI-driven attacks, increasingly targets Higher Education Institutions (HEIs) through e-learning systems. This study introduces the AI-SEC-EDU framework to guide the integration of security controls and AI-enabled intelligence into cybersecurity strategies for e-learning platforms. The framework is based on a narrative review of existing cybersecurity interventions for e-learning in Low-Income Countries (LICs) and their approach to managing cybersecurity in the age of Artificial Intelligence. A search across four databases—ACM, Springer, ScienceDirect, and Google Scholar—in May 2025 identified 621 papers, of which eight met the inclusion criteria using PICO and PRISMA guidelines. The selected papers focused on cybersecurity in e-learning, discussing frameworks, models, and algorithms for platforms like Moodle, Google Classroom, and Coursera, some of which incorporate AI and open-source options. The study identifies three key security risk domains: technological infrastructure, human factors, and institutional governance, all of which are compounded by limited AI integration. Existing measures focus on system hardening but fail to address AI-based threat prediction and human behavior vulnerabilities. The AI-SEC model integrates AI, user awareness, and governance controls to provide adaptive, context-sensitive cybersecurity solutions for e-learning in LICs. This framework serves as a diagnostic and planning tool, aligning policies, institutional practices, and national strategies.

A Lightweight ITSM Framework for Balancing Service Value and Cost Efficiency in Digital MSMEs

Marcel Marcel, Jessica Alexander — Tue, 09 Dec 2025 20:10:26 +0700

This study explores how Indonesian Micro, Small, and Medium Enterprises (MSMEs) manage IT services with limited staff and budgets, proposing that a simpler, lightweight approach to IT service management (ITSM) is more suitable for this context. A mixed-method case study was conducted with four MSMEs in Greater Jakarta—two technology-based and two non-technology-based. Key performance indicators such as response time, downtime, and customer satisfaction were derived from service logs and customer ratings, while semi-structured interviews with owners and staff were analyzed for recurring themes. Results revealed that non-technology-based MSMEs achieved a median response time of 12.5 minutes and an average satisfaction score of 4.55, while technology-based MSMEs had a median response time of 1.8 hours and an average score of 3.95. Technology firms logged approximately seven hours of downtime per month, compared to 1.5 hours in non-tech firms, indicating a trade-off between faster responses and higher satisfaction at the cost of less systematic documentation and control. All MSMEs utilized freemium SaaS tools, marketplace dashboards, limited-service hours, and no dedicated IT staff to minimize costs. The study proposes a lightweight ITSM framework and checklist, adaptable with free tools, for use in MSME incubators and support programs, advancing ITSM literature for resource-constrained businesses.

Enhancing Smart Wheelchair Control: A Comparative Study of Optical Flow and Haar Cascade for Head Movement

Nimatul Ma Muriyah, Paerin Paerin, Andik Yulianto — Tue, 09 Dec 2025 20:35:15 +0700

The development of Artificial Intelligence, particularly in Computer Vision, has enabled real-time recognition of human movements such as head gestures, which can be utilized in smart wheelchairs for users with limited mobility. This study compares two lightweight non-deep-learning methods Lucas–Kanade Optical Flow and Haar Cascade Classifier for real-time head movement detection. Both methods were implemented in Python using OpenCV and tested in four basic directions (left, right, up, and down) under three different lighting conditions: bright, normal, and dim. Each condition consisted of 16 trials per method, resulting in a total of 96 trials. The evaluation focused on detection accuracy and decision time. Under bright lighting, Optical Flow achieved 87.5% accuracy with a decision time of 0.338-1.41 s, while Haar Cascade reached 50% accuracy with 0.616–1.20 s. Under normal lighting, Optical Flow maintained 87.5% accuracy with 0.89–1.21 s, compared to Haar Cascade’s 68.75% accuracy with 0.83–1.25 s. Under dim lighting, Optical Flow improved to 93.8% accuracy with 0.90–1.31 s, whereas Haar Cascade dropped to 62.5% accuracy with 0.89–1.58 s. These findings confirm that Optical Flow delivers more reliable and adaptive performance across varying illumination levels, making it more suitable for real-time smart wheelchair control. This study contributes to the development of affordable assistive technologies and highlights future directions for multi-user testing and hardware integration.

Sentiment Classification of TikTok Reviews on Almaz Fried Chicken Using IndoBERT and Random Oversampling

Imam Syahputra Zaki, Rizka Dhini Kurnia, Allsela Meiriza — Tue, 09 Dec 2025 21:35:56 +0700

The socio-political context surrounding the Indonesian Ulema Council's Fatwa No. 83 of 2023, which catalyzed a significant consumer shift, necessitates an accurate measure of public sentiment toward alternative local brands like Almaz Fried Chicken. Analyzing real-time consumer discourse on the challenging TikTok platform, the study utilized a final dataset of 4,374 unique comments to overcome the inherent problem of dataset imbalance and linguistic informality. The core method involved a seven-stage quantitative approach: data collection, preprocessing, sentiment labeling, data splitting (70:15:15), Random Oversampling (ROS), IndoBERT fine-tuning, and evaluation. This pipeline fine-tuned IndoBERT, a Transformer-based model, integrated with ROS applied exclusively to the training data. Evaluation demonstrated that ROS significantly reduced model bias and enhanced performance: Overall Accuracy increased by 2.0% (from 91% to 93%), and the Macro F1-Score improved by 3.4% (from 0.87 to 0.90). Most critically, the F1-Score for the minority Negative sentiment class surged from 0.78 to 0.84, confirming ROS's effectiveness in accurately detecting critical feedback. These findings provide timely, data-driven insights into brand perception amidst the boycott campaign and establish a robust, reliable IndoBERT-ROS methodology for advanced sentiment monitoring in dynamic social media environments.

Trends of Machine Learning, Cybersecurity and Big Data Analytics in Industry 4.0

Tue, 09 Dec 2025 22:42:20 +0700

This research explores the integration of Machine Learning (ML), Cybersecurity, and Big Data Analytics (BDA) in advancing intelligent, secure, and sustainable industrial ecosystems within Industry 4.0. It assesses global research productivity, collaboration patterns, and the connection between intelligent automation, data-driven innovation, and cyber resilience. A PRISMA-based bibliometric review of 1,386 relevant publications from the Scopus database (2020-2025) was conducted, using Biblioshiny visualization tools to map key authors, institutions, countries, and emerging research clusters. Findings show a 7.09% annual growth in publications, reflecting a growing global focus on ML, BDA, and cybersecurity within Industry 4.0 ecosystems. The United States, China, and India were identified as major contributors, with strong cross-continental collaborations fostering innovation. Key research topics include deep learning, digital twins, and the Internet of Things (IoT), while emerging areas such as explainable AI, federated analytics, and edge computing are gaining attention. By mapping global research dynamics and identifying key contributors, this study highlights critical research gaps and offers practical insights for advancing interdisciplinary innovation, aimed at creating secure, intelligent, and sustainable industrial ecosystems in Industry 4.0.

Predicting Accounts Receivable of the Social Security Administration for Employment Using LSTM Algorithm

Ainna Khansa, Usman Ependi — Tue, 09 Dec 2025 23:12:19 +0700

This study explores the use of Long Short-Term Memory (LSTM) networks for predicting outstanding contributions from employers to the BPJS Ketenagakerjaan, Indonesia’s social security agency. The research aims to address the challenges BPJS faces due to delayed or unpaid contributions, which impact the institution's operational stability and financial health. The LSTM model, a deep learning technique well-suited for time-series prediction, was applied to historical data from BPJS Ketenagakerjaan to predict overdue contributions across three different training-validation splits: 70:30, 80:20, and 90:10. The results demonstrate that the 80:20 split achieved the highest validation accuracy of 84.71%, offering the optimal balance between training data and model generalization. The model's ability to predict overdue contributions with high accuracy could significantly improve BPJS's receivables management, allowing for more proactive financial planning and risk mitigation. The study also highlights the integration of an attention mechanism within the LSTM model, enhancing its predictive capabilities by focusing on the most relevant historical data. This research contributes to the field of predictive analytics in public sector financial management, showcasing the potential of machine learning in enhancing the efficiency and effectiveness of social security programs.

Web Information System for E-Sport Arena Community with OWASP-Based Cybersecurity Using XP Method

Afgha Aufan, Purwadi Purwadi, Argiyan Dwi Pritama — Wed, 10 Dec 2025 14:17:57 +0700

The rapid development of the e-sports ecosystem in Purwokerto has encouraged the emergence of digital communities such as the Esport Arena Community. However, the management of member data, event information, and merchandise transactions is still carried out manually through social media, resulting in inefficiency and limited-service reliability. This study aims to develop a web-based information system that integrates member management, jersey evolution documentation, and secure online merchandise transactions. The system was developed using the Extreme Programming (XP) method, which supports iterative development and continuous refinement. Security measures were implemented based on OWASP Top 10 recommendations, including prepared statements, input validation, CSRF protection, and Role-Based Access Control (RBAC). System evaluation using the System Usability Scale (SUS) produced a score of 88, categorized as Excellent, indicating high user satisfaction and strong usability performance. The results demonstrate that the system operates securely, reliably, and effectively improves operational efficiency for the Esport Arena Community.

Stock Price Prediction Using Backpropagation ANN: Case Study of ADMR (2023–2025)

Muhammad Khozin, Zainal Abidin, Totok Chamidy — Wed, 10 Dec 2025 14:53:13 +0700

This study develops an Artificial Neural Network (ANN) backpropagation model for predicting stock prices using ADMR stock data from 2023 to 2025, obtained from Yahoo Finance. Given the inherent volatility and unpredictability of stock prices, accurate forecasting plays a crucial role in investment decision-making. ANN models are particularly effective for capturing complex, non-linear relationships and patterns in financial data, which traditional statistical models may fail to address. In this research, various configurations were tested by adjusting the number of hidden neurons (5, 10, and 15) and learning rates (0.1, 0.3, and 0.5). The optimal model architecture was found to be 3-10-1, consisting of three input neurons, ten hidden neurons, and one output neuron, which achieved the best prediction performance with a Mean Absolute Percentage Error (MAPE) of 2.26%. This model was trained with a learning rate of 0.3 and completed in 915 iterations. However, the model's predictive capabilities are constrained by its reliance on historical stock prices alone, excluding external factors such as macroeconomic indicators, market sentiment, or trading volume, which may improve its generalization and overall accuracy. Future work could integrate these variables for better robustness and predictive power.

Utilizing Random Forest Method for Predicting Student Dropout Risk in Madrasah Environments

Muhammad Mahsun, M. Amin Hariyadi, Sri Harini — Wed, 10 Dec 2025 16:43:39 +0700

The phenomenon of school dropout represents a crucial issue with negative impacts on educational institution performance, social stability, and national development. Consequently, the early detection of high-risk students constitutes a strategic preventive measure. This research aims to develop an accurate predictive model using a Machine Learning approach. The study employed a comparative evaluation using classification algorithms, with the primary focus being the performance analysis of the Random Forest Classifier. The dataset utilized, comprising 1,763 student records, underwent a rigorous data pre-processing phase, including data cleaning, variable transformation, and class imbalance handling, to ensure high-quality input. The model was trained using a Random Seed configuration of 75 to guarantee experimental reproducibility and consistency in evaluation results. Experimental findings indicate that the Random Forest algorithm provided the best performance, achieving an accuracy of 82.0% and a precision of 83.8%. This superior performance confirms the model's effectiveness in identifying the key determinants of dropout, stemming from both students' internal and external factors. Based on these results, the research recommends the application of Random Forest as a Decision Support System instrument to facilitate targeted interventions, including medical support, economic assistance, and academic counseling. Future research is advised to integrate historical counseling data to further enhance the prediction sensitivity of the model.

Taxpayer Classification Using K-Means Clustering to Support CRM Strategy Development: Case Study of Prabumulih City Samsat

Bimmo Fathin Tammam, Ali Ibrahim, Dwi Rosa Indah, Ahmad Fali Oklilas, Yadi Utama — Wed, 10 Dec 2025 20:45:57 +0700

Effective management of taxpayer data is crucial for enhancing compliance and optimizing regional revenue. This study addresses the limited use of data-driven taxpayer segmentation in local Samsat institutions by applying K-Means Clustering to support targeted Customer Relationship Management (CRM) strategies. A dataset of 3,999 motor vehicle taxpayer records from September 2025 was processed through feature selection, scaling, and clustering. The analysis identified three distinct taxpayer groups based on payment timeliness, compliance consistency, and vehicle age. Cluster validity was confirmed using the Davies-Bouldin Index, yielding a value of -41.327 for k = 3, supported by ANOVA for statistical significance. The findings highlight how clustering can reveal taxpayer behavior patterns, guiding personalized services and compliance programs. This study's novelty lies in integrating clustering outcomes with practical CRM strategies for public agencies, offering a data-driven approach to improve taxpayer engagement and regional revenue. However, the study is limited by its focus on a single-period dataset and vehicle-related attributes.

Hybrid Random Forest Regression and Ant Colony Optimization for Delivery Route Optimization

Reni Aurelia, Abdul Rahman — Wed, 10 Dec 2025 21:21:50 +0700

The transportation of goods in Indonesian cities is increasingly challenged by urbanization, congestion, diverse road characteristics, and environmental factors, reducing the effectiveness of conventional distance-based routing. This study enhances delivery route optimization by integrating travel-time prediction using Random Forest Regression (RFR) with a metaheuristic routing process using Ant Colony Optimization (ACO). Using OpenStreetMap (OSM) data for Palembang, experiments were conducted on five simulated customer locations in Zone 1. Road attributes such as segment length, road type, and estimated speed were used to train the RFR model, whose predicted travel times served as dynamic costs in the ACO heuristic. The RFR model achieved high predictive accuracy (R² = 0.98; MSE = 8.81), and the ACO-based optimization produced an efficient route of 29.58 km with a total travel time of 148 minutes. However, the experiment is limited to a single zone, a small number of customers, and the removal of real traffic variables—where all actual speed variations, congestion levels, and time-dependent traffic conditions were simplified or omitted, causing the model to rely solely on static road attributes. Future work will incorporate real-time traffic data, expand testing to multiple zones, and use larger datasets to improve scalability and operational applicability.

Multivariate LSTM for Drug Purchase Prediction in Pharmaceutical Management

Fanny Brawijaya, Agung Teguh Wibowo Almais, Totok Chamidy — Wed, 10 Dec 2025 00:00:00 +0700

This study aims to develop a structured approach to predict the number of hospital drug purchases using deep learning techniques. The Multivariate Long Short-Term Memory (LSTM) model is designed to capture temporal and contextual patterns including transaction time, polyclinic type, and drug type to improve the efficiency of pharmaceutical management. The model was tested using outpatient transaction data at RSIA Fatimah Probolinggo hospital in East Java, Indonesia, through three configurations (A, B, and C) to determine the optimal parameters. The best model, the Model B1, produces a Mean Absolute Error (MAE) value of 10.239, Mean Absolute Percentage Error (MAPE) of 1.976%, and the Coefficient of Determination (R²) of 0.199, which indicates a high degree of accuracy. The results of the study prove that multivariate LSTM is able to model complex intervariable dependencies and provide superior results than conventional forecasting methods. In practical terms, this model can be used as a decision-making tool for hospital management in planning drug procurement, optimizing inventory, and preventing shortages and overstocks. The application of this model contributes to data-driven pharmaceutical supply chain planning in smart hospital management systems.

Integrating Human-Centered AI into the Technology Acceptance Model: Understanding AI-Chatbot Adoption in Higher Education

Fine Masimba, Kudakwashe Maguraushe, Bester Chimbo — Wed, 10 Dec 2025 00:00:00 +0700

Artificial intelligence (AI) is transforming education by enhancing assessments, personalizing learning, and improving administrative efficiency. However, the adoption of AI-powered chatbots in higher education remains limited, primarily due to concerns about trust, transparency, explainability, perceived control, and alignment with human values. While the Technology Acceptance Model (TAM) is commonly used to explain technology adoption, it does not fully address the challenges posed by AI systems, which require human-centered safeguards. To address this gap, this study extends TAM by incorporating Human-Centered AI (HCAI) principles—explainability, transparency, trust, and perceived control—resulting in the HCAI-TAM framework. An empirical study with 300 respondents was conducted using a structured English questionnaire, and regression analysis was applied to assess the relationships among variables. The model explained 65% (R² = 0.65) of the variance in behavioral intention and 55% (R² = 0.55) in usage behavior. The findings highlight that integrating HCAI principles into TAM enhances user adoption of AI chatbots in higher education, contributing both theoretically and practically.

IoT-Based Smart Fertigation System for Citrus Plants using Fuzzy Logic Control

Mon, 10 Nov 2025 00:00:00 +0700

This study develops and evaluates an IoT-based smart fertigation system for citrus plants using a fuzzy logic control (FLC) algorithm integrated with an ESP32 microcontroller and wireless sensors. The research aims to address the limitations of conventional fertigation practices that rely on fixed schedules without considering real-time soil and climate conditions, which often result in water inefficiency and nutrient imbalance. The developed system integrates sensors for soil moisture, temperature, and air humidity to automatically regulate irrigation duration through triangular membership functions and a fuzzy inference model consisting of 64 fuzzy rule combinations. Over a 30-day observation period, twelve citrus seedlings aged 3-4 months after grafting were organized into four experimental groups: manual fertigation (MF), manual irrigation (MI), smart fertigation (SF), and smart irrigation (SI). Experimental findings indicated that the smart fertigation system-maintained soil moisture within a more stable range of 40-55%, and plants in the SF group experienced approximately 7 cm greater height and a twofold increase in leaf count compared to manually irrigated ones. The smart fertigation treatment also produced more uniform, greener, and healthier foliage, signifying balanced nutrient uptake. Overall, the IoT–FLC integration provides a more adaptive and eco-efficient irrigation model that promotes sustainable management of water and nutrients in tropical citrus cultivation.

Child Nutrition Prediction for Stunting Prevention Using K-Nearest Neighbor (K-NN) Algorithm

Chandra Kirana, Delpiah Wahyuningsih, Louis Michael — Wed, 10 Dec 2025 00:00:00 +0700

Stunting is a significant public health issue in Indonesia, affecting both children's physical growth and cognitive development. This study aims to develop a child nutritional status prediction application using the K-Nearest Neighbor (K-NN) algorithm as an early detection tool for stunting prevention. The model classifies nutritional status into five categories: good nutrition, poor nutrition, undernutrition, overnutrition, and obesity, using anthropometric data such as age, weight, height, and gender. The dataset comprises 49,766 samples of children aged 0–5 years from the Bangka Belitung Islands Provincial Health Office. The data processing included normalization, feature selection, and k-value testing to optimize model performance. Evaluation results showed that K-NN with k = 2 achieved 92% accuracy, with the best precision and recall in the good nutrition category (0.94 and 0.99). However, performance in minority categories like malnutrition remains low due to data imbalance. The weighted averages for precision, recall, and F1-score were 0.90, 0.92, and 0.90, respectively. This research's novelty lies in integrating the K-NN model into a mobile application, enabling real-time nutritional status assessment for health workers, improving fieldwork efficiency, and facilitating early detection and monitoring.

Agile Methodologies as Drivers of Organizational Culture in Digital Transformation Projects

Tendamudzimu Murudi, Thulani Lucas Khoza — Wed, 10 Dec 2025 00:00:00 +0700

Organisations are increasingly adopting agile methodologies to accelerate digital transformation, yet outcomes remain inconsistent when agile is viewed solely as a delivery technique rather than a cultural mechanism. This study explores how agile practices influence organisational culture during transformation and identifies the conditions under which benefits are realised. Using an explanatory mixed-methods design, the study combined a survey (N = 315) with 18 semi-structured interviews. After ensuring scale reliability, bivariate correlations and linear/multiple regressions were conducted in IBM SPSS, while interview transcripts were thematically coded to explain the quantitative findings. Agile practices were positively linked to a supportive culture (Pearson r ≈ .32), with simple regression indicating that agile significantly predicted culture (R² ≈ 0.10). A multiple regression predicting digital-transformation outcomes from agile and culture revealed a significant model (R² ≈ 0.40), where agile was the stronger predictor and culture made a smaller yet meaningful contribution. Qualitative insights highlighted how cadence, visibility, and iteration normalised collaboration, transparency, and rapid feedback. Practically, managers should focus on culture outcomes in agile roadmaps, institutionalise essential routines, and reduce structural hand-offs for sustained transformation.

Optimized K-Means Clustering for Web Server Anomaly Detection Using Elbow Method and Security-Rule Enhancements

Rahmawan Bagus Trianto, Muhammad Abdul Muin, Cahya Vikasari — Wed, 10 Dec 2025 00:00:00 +0700

Anomaly detection in web server environments is essential for identifying early indicators of cyberattacks that arise from abnormal request behaviors. Traditional signature-based mechanisms often fail to detect emerging or obfuscated threats, requiring more adaptive analytical approaches. This study proposes an optimized anomaly detection model using K-Means clustering enhanced with engineered security-rule features and the Elbow Method. Two datasets were used: a small dataset of 3,399 log entries from one VPS and a large dataset of 223,554 entries collected from three VPS nodes, all sourced from local production servers of the Department of Computer and Business, Politeknik Negeri Cilacap. The preprocessing pipeline includes timestamp normalization, removal of non-informative static resources, numerical feature scaling, and TF-IDF encoding of URL paths. Domain-driven security features entropy scores, encoded-payload indicators, abnormal status-code ratios, and request-rate deviations were integrated to improve anomaly separability. Experiments across five model configurations show that combining larger datasets with rule-based features significantly enhances clustering performance, achieving a Silhouette Score of 0.9136 and a Davies–Bouldin Index of 0.4712. The results validate the effectiveness of incorporating security-rule engineering with unsupervised learning to support early-warning threat detection in web server environments.

Utilizing ROP in a Mobile Based Warehouse Management System for Small Retail Businesses

Winni Setiawati, Jap Tji Beng, Tony Tony — Wed, 10 Dec 2025 00:00:00 +0700

ABC Store, a wholesale sock retailer, operates through both physical and online platforms. With increasing online sales, the company has experienced frequent stockouts due to inefficient manual inventory processes. This study aims to address these challenges by developing a mobile-based Warehouse Management System (WMS) integrated with a Reorder Point (ROP) calculation. The system was developed using the V-Shaped model, which involved requirement gathering through interviews and warehouse observations. The application was built using the lightweight Flutter framework and an SQLite local database, with testing including white-box, black-box, and User Acceptance Testing (UAT). Additionally, QR code scanning was implemented to improve document tracking and inventory management. The results indicate that the system functions as expected, with the ROP calculation effectively supporting restocking decisions and minimizing stockouts. This study contributes to the field by demonstrating the practical application of integrating simplified ROP calculation into a mobile-based WMS, highlighting its potential to improve warehouse operations for small businesses with limited infrastructure. The approach offers a scalable solution for managing inventory efficiently and cost-effectively, especially in small-scale retail settings.

The IoT-Based E-Voting System Using Fingerprint Biometrics for School Elections

Wahyu Andre Wibowo, Erik Iman Heri Ujianto — Wed, 10 Dec 2025 00:00:00 +0700

This study proposes an Internet of Things (IoT)-based e-voting system to address the limitations of traditional paper-based student council elections, which are prone to errors, inefficiency, and data manipulation. The system is developed using the ADDIE model for Research and Development (R&D), incorporating a Laravel-based administrative dashboard, a Flutter-based mobile voting interface, and a biometric authentication device built with an ESP32 microcontroller and JM-101B fingerprint sensor. Evaluation involved 20 participants who completed six functional test scenarios, achieving a 100% success rate across 120 instances. Usability testing revealed a notable comfort difference, with 30% comfort on mobile phones and 90% on tablets. Performance testing showed a fingerprint scan time of 669.6 ms and a vote submission latency of 437.1 ms, indicating good system responsiveness. The results suggest the system improves security, transparency, and efficiency in the election process. However, the study is limited by a small sample size and evaluation within a single institution. Future work could explore cloud integration, multi-school deployment, and additional authentication methods to enhance scalability and support broader adoption.

Quantum Computing Cryptography: A Systematic Review of Innovations, Applications, Challenges, and Algorithms

Peter Maitireni, Vusumuzi Ncube, Belinda Ndlovu, Thando Sibanda — Wed, 10 Dec 2025 00:00:00 +0700

This study explores how to build quantum-resistant systems to safeguard digital infrastructure in the post-quantum era by uncovering the innovations, applications, algorithms, and challenges of Quantum Computing cryptography. Utilizing the Preferred Reporting Items for Systematic Reviews and Meta-Analyses approach a search was conducted across the following databases for the years 2021–2025: PubMed, IEEE Xplore, ScienceDirect, SpringerLink, and Google Scholar. We shortlisted 15 studies from 519 screened articles for a comprehensive evaluation based on their relevance. Findings show strong adoption in finance, healthcare, IoT, cybersecurity, and e-government, with lattice-based PQC emerging as the most dominant cryptographic family, followed by QKD and hybrid PQC–QKD models. The review highlights key obstacles, including transition complexity, lack of global standards, high implementation costs, and integration difficulty. The study contributes by providing the first sector-aligned synthesis of innovations, identifying algorithmic trends, and mapping global research disparities through a conceptual model. It also presents a structured set of future research directions to guide policymakers, cryptographers, and practitioners preparing for quantum-enabled threats.

Adaptive-Delta ADWIN: A Framework for Stable and Sensitive Intrusion Detection in Streaming Networks

Rodney Buang Sebopelo — Fri, 12 Dec 2025 11:07:42 +0700

Intrusion Detection Systems (IDS) must adapt to network traffic streams where concept drift alters the normal and malicious behaviors. The traditional drift detectors with fixed sensitivity parameter () fail to balance responsiveness and stability, reducing detection reliability. This study introduces Adaptive−Delta ADWIN framework that adjusts through two online controllersthe Volatility Controller (VC) and AlertRate Controller (ARC) to improve the sensitivity while maintaining stability. The experiments were evaluated on the CICIDS2017 dataset using multiclass ensemble of Hoeffding Adaptive Trees, the framework achieved 0.930.95, surpassing fixed baselines by up to 6.6%. The false positive and false negative rates were reduced by 50% and 30%. Overall, the results confirm that Adaptive ADWIN enhances multiclass IDS performance between detection sensitivity and operational stability in the realtime network conditions.

K-Means Clustering with Elbow Method for Stunting Risk Detection in Toddlers Using Anthropometric and Nutritional Data

Fri, 12 Dec 2025 11:43:56 +0700

Stunting remains a critical public health challenge in Indonesia, primarily due to inadequate nutrition and recurrent infections in early childhood. This study aimed to identify patterns of stunting risk by integrating anthropometric and dietary data, specifically sugar consumption, using an unsupervised machine learning approach. A total of 20 toddlers aged 12-59 months from Purwokerto Selatan participated. Anthropometric data (age, weight, height) and dietary intake (sugar consumption, snack frequency) were collected via a caregiver questionnaire. K-Means clustering was applied, with the optimal number of clusters determined using the Elbow Method (K=2). Two clusters were identified: Cluster 0, with a lower risk of stunting, and Cluster 1, with a higher proportion of toddlers at risk. Cross-tabulation with stunting status validated this, showing that Cluster 1 contained more children with "Potential" stunting. Internal validation using the Silhouette score (0.252) and PCA visualization confirmed the clustering's robustness. This study demonstrates the potential of combining anthropometric and dietary data for stunting risk profiling, suggesting a complementary approach for growth monitoring programs and targeted interventions.

Stakeholder Analysis for Enhancing Ethics in Software Development: A Scoping Review

Senyeki Milton Marebane, Ernest Mnkandla — Fri, 12 Dec 2025 13:57:41 +0700

Stakeholder analysis has become a crucial means for achieving software development ethical goals. This scoping review study aims to provide an overview of the extent to which stakeholder analysis has been applied to enhance ethics in software development. The study explores research studies that have been published by IEEE, ACM, Science Direct, AIS and Journal of Business Ethics. PRISMA-ScR was employed to achieve the objective of the study. Only six research studies published from 1999 to 2015 met the selection criteria for data extraction and analysis. The results show that the focus of stakeholder analysis is on moral impulse, risk identification, stakeholder interactions, stakeholder classification and impact level. Furthermore, stakeholder analysis for enhancing ethics in software development is prevalent to empower the voiceless stakeholders, risk management and stakeholder mapping and quantification for measuring stakeholder impact on projects. The analysis of the results reveals several research gaps such as unavailable empirical studies beyond 2015, concentration only on requirements engineering and lack of studies on stakeholder analysis on emergent technologies. The implications of this study point to the need for more guidance and expanded use of stakeholder analysis across the complete software process to benefit software teams and ongoing research to harness the potential of this theory on enhancing ethics in the emergent technologies.

AIDA-Based Analysis of TikTok Live Marketing at Bin Dawood Boutique Purwokerto

Revalyna Octavia Maharani, Argiyan Dwi Pritama, Luzi Dwi Oktaviana — Fri, 12 Dec 2025 20:13:48 +0700

This study aims to analyze the TikTok Live strategy implemented by Bin Dawood Butik Purwokerto using the AIDA model as an analytical framework. This research employed a descriptive qualitative design, through observations of 10 TikTok Live sessions, in-depth interviews with a marketing administrator and a live host, and documentation. Data were analyzed through reduction, presentation, and conclusion drawing based on the AIDA framework, with triangulation applied to ensure validity. Bin Dawood Boutique’s TikTok Live strategy effectively implements the AIDA model. The attention stage is built through persuasive greetings, engaging visuals, and broadcast scheduling that aligns with audience activity patterns. The interest stage is strengthened through the use of TikTok’s interactive features and the delivery of clear, responsive product information. The desire stage develops through host transparency, emphasizing product benefits, and strategic urgency. The action stage is driven by the use of clear CTAs and seasonal factors, such as the Eid al-Fitr period and payday, which encourage consumers to make transactions. Cumulative TikTok sales reached more than 26,900 products, confirming TikTok Live’s significant contribution to overall sales performance. This study provides an empirical overview of how a local fashion business operationalizes each stage of the AIDA model on TikTok Live. As a limitation, the study focuses on a single case and relies on qualitative data, which may constrain generalization.

Emerging Trends in Digital Transformation and Information Systems by Bibliometric Analysis in the United States

Sat, 13 Dec 2025 22:40:13 +0700

This bibliometric analysis examines the evolving trends of digital transformation (DT) and information systems (IS) in U.S. businesses. It explores how technology—focused on strategy, efficiency, innovation, and customer engagement—is reshaping organizations and workplaces. Using a PRISMA-based systematic review and data from Scopus (2016–2026), the study applies the Bibliometrix R package to assess publication patterns. Results show significant growth, with 2,692 documents reflecting a 43.58% annual increase and 18.39% involving international collaboration. Key themes include AI/ML integration in business processes, digital sustainability, and IS as a strategic driver for business model evolution. U.S. businesses are increasingly aligning digital transformation with sustainability goals. This study addresses a key research gap by offering detailed insights into DT and IS impacts on operations and sustainability practices. It underscores the need for integrated socio-technical strategies, responsible data governance, and global collaboration to foster innovation and bridge digital divides.

Multi-Tier Architecture Design for Scalable and Effective Non-Formal Learning: A Redesign of Serat Kartini Women's School LMS

Primavieri Rhesa Ardana, Gustina Alfa Trisnapradika — Sat, 13 Dec 2025 23:14:54 +0700

Non-formal education plays a vital role in empowering women in rural areas of Central Java, Indonesia. However, the existing Learning Management System (LMS) of Woman School Serat Kartini, built on a monolithic Laravel architecture, suffers from significant performance degradation and scalability limitations under growing user loads and shared hosting constraints. This leads to high latency, frequent session interruptions, and reduced participation, ultimately undermining learning effectiveness. This study redesigns the LMS using a multi-tier application architecture through the Design Science Research (DSR) methodology. The proposed blueprint separates the system into four independent tiers: Presentation (Next.js for users, React.js for administrators), Logic (Express.js for API Layer), Cache (Redis with cache-aside strategy), and Data (MySQL). The design artifacts include detailed architecture diagrams, ERD, use case, and sequence diagrams. Conceptual evaluation demonstrates that the multi-tier approach enhances modularity, reduces latency, supports horizontal scalability, and improves resource efficiency , ensuring reliable access for women learners with limited digital literacy and unstable internet connectivity. The redesigned LMS conceptually strengthens learning accessibility, engagement, and program sustainability in resource-constrained non-formal education contexts. This research is limited to the conceptual design phase without implementation or empirical testing.

Sentiment Analysis of User Reviews for the PLN Mobile Application Using Naïve Bayes and Long Short-Term Memory

Sat, 13 Dec 2025 00:00:00 +0700

This study explores large-scale sentiment analysis of user reviews for the PLN Mobile application to better understand public perception and provide quantitative insights for improving digital electricity services in Indonesia. Addressing the lack of benchmarks for Indonesian public-service apps—where prior studies rely on smaller datasets and traditional machine learning—this research positions sentiment analysis as a tool for continuous user experience monitoring. A total of 50,000 Indonesian-language reviews from Google Play were collected and pre-processed using cleaning, case folding, tokenization, stopword removal, normalization, and stemming. Sentiments (positive, neutral, negative) were assigned using a domain-specific Indonesian sentiment lexicon, yielding approximately 40% positive, 35% neutral, and 25% negative labels. Two models were applied: Multinomial Naïve Bayes using TF-IDF features and a Long Short-Term Memory (LSTM) model with 100-dimensional word embeddings and a 128-unit LSTM layer. Naïve Bayes achieved 70.89% accuracy (F1-score: 0.6964), while LSTM outperformed it with 98.02% accuracy (F1-score: 0.9800). These results highlight the superiority of deep learning in sentiment monitoring and offer a scalable framework to help PLN and policymakers enhance digital public service delivery.

Cybersecurity Trends in Digital Marketing for Public Health: A PRISMA based Bibliometric Analysis

Sat, 13 Dec 2025 00:00:00 +0700

This study conducts a bibliometric analysis of digital marketing in public health through the lens of cybersecurity, aiming to evaluate research trends from 2015 to 2025. It identifies key developments, major contributors, and provides guidance for future studies. A total of 1,191 documents were analyzed, revealing a significant annual growth rate of 32.01% and an average of 19.01 citations per document. The analysis explores how digital marketing for public health intersects with cybersecurity, a domain that remains underexplored. Data collection and visualization were conducted using Scopus, Biblioshiny, and VOSviewer, with article selection guided by the PRISMA methodology. Results indicate consistent growth in publications over the decade, though a noticeable decline occurred post-COVID-19 in 2020. The study offers a comprehensive mapping of existing literature and highlights the strategic importance of integrating cybersecurity into digital health marketing to protect patient data, maintain public trust, and enhance health outcomes. It provides valuable insights for researchers, policymakers, and practitioners aiming to improve the security and effectiveness of digital health communication in an increasingly connected world.

Oil and Gas Production Forecasting Based on LSTM Model: A Case Study of PT Pertamina Hulu Rokan Zone 4

Angel Caroline Billan, Usman Ependi — Mon, 15 Dec 2025 11:10:42 +0700

This study addresses the critical need for accurate oil and gas production forecasting to support strategic decision-making in Indonesia’s energy sector. PT Pertamina Hulu Rokan Zone 4 (PHR Zona 4), a key player in national energy production, frequently encounters technical and external operational challenges. To tackle these issues, this research proposes a deep learning-based predictive model using the Long Short-Term Memory (LSTM) architecture, structured in an encoder-decoder format and enhanced with an attention mechanism. The model was trained and tested on historical oil and gas production data from PHR Zona 4, evaluated under two data-splitting scenarios: 80:20 and 90:10. Model performance was assessed using Mean Absolute Percentage Error (MAPE), Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and the coefficient of determination (R²). Results from the 80:20 scenario showed RMSE of 5.83, MAE of 5.54, MAPE of 1.71%, and R² of -1.97, suggesting difficulties in capturing extreme data fluctuations. However, the 90:10 scenario demonstrated significantly improved performance with RMSE of 0.42, MAE of 0.36, MAPE of 0.11%, and R² of 0.00, indicating better trend prediction stability. The novelty of this study lies in the integration of attention mechanisms within the LSTM encoder-decoder framework for oil and gas time series forecasting, offering enhanced accuracy and robustness. This research provides a valuable foundation for future improvements in predictive analytics and operational efficiency in the oil and gas industry.