Jurnal ICT : Information and Communication Technologies

Data Mining for Drug Inventory Using Web-Based FP-Growth Method

Della Puspita, Wirhan Fahrozi — Sun, 26 Oct 2025 00:00:00 +0000

Pharmacy inventory management plays a critical role in ensuring product availability and preventing financial losses due to overstock or stock shortages. However, many pharmacies, including Global Medcare Pharmacy in Medan, still experience challenges in maintaining optimal inventory levels due to the absence of data-driven management systems. This study aims to develop and implement a website-based data mining application using the FP-Growth algorithm to identify frequent itemsets and uncover association patterns within pharmaceutical sales transactions. The FP-Growth algorithm was applied to 300 transaction records to generate frequent item combinations with a minimum support threshold of 3%. The results reveal strong associations among specific drugs, such as Amlodipine 10mg, Azithromycin 500mg, and Cetirizine 10mg, with confidence levels reaching up to 100%. These findings demonstrate that FP-Growth effectively identifies purchasing patterns that can guide pharmacies in forecasting demand, managing stock levels, and designing promotional bundles. The practical implication of this research is that integrating FP-Growth into pharmacy information systems can enhance decision-making accuracy, improve service quality, and increase operational efficiency. Nevertheless, the study is limited to a single-site dataset and static analysis; future research should employ larger datasets and hybrid predictive approaches for real-time implementation across multiple pharmacy networks.

Implementation of Augmented Reality in Foreign Languages (English - Arabic) Using the Marker-Based Tracking Method

Joel Dakosta Banjarnahor, Nita Syahputri — Fri, 31 Oct 2025 00:00:00 +0000

Mastery of foreign languages, particularly English and Arabic, remains a challenge for Indonesian students due to limited exposure, low motivation, and the inadequacy of conventional learning media. This study aims to enhance students’ interest and learning outcomes in English and Arabic through the development and implementation of a Marker-Based Augmented Reality (AR) learning application. The research employs a design and development methodology, integrating AR with marker tracking using Vuforia and Unity, and evaluates the application across key functional scenes, including vocabulary recognition, interactive 3D object projection, quizzes, and learning materials. The results indicate that the AR application significantly improves engagement, motivation, and comprehension, allowing students to experience contextualized learning in an interactive and flexible environment. Marker-Based Tracking ensures accurate real-time projection of virtual objects onto physical markers, supporting multi-modal language acquisition, including vocabulary, pronunciation, and contextual understanding. Furthermore, the application demonstrates practical usability, scalability, and adaptability for self-directed and blended learning, highlighting its potential as a complementary educational tool. These findings suggest that AR can serve as an effective innovation to address low interest and achievement in foreign language learning, providing a model for immersive and interactive pedagogical practices. However, limitations such as device dependency and short-term evaluation underscore the need for further research to examine long-term learning outcomes, wider population applicability, and integration with markerless AR technologies. Overall, the study contributes to the emerging field of AR-mediated language learning by offering both theoretical insights and practical guidance for educators and developers.

Designing a Website-Based Study Program Management Application and Digital Attendance for Informatics Study Programs

Thu, 30 Oct 2025 00:00:00 +0000

Conventional attendance and conventional management of academic activities often cause problems, such as recording errors, delays, and lack of access to organized information. In addition, academic communication still relies on external platforms, such as Telegram, which does not support the needs of lecture activity management. This research aims to design a web-based application called ClassTIcs, which provides digital attendance features, lecture schedule management, lecture material upload, and textbased discussion forums. This application involves three main actors: staff as super admin, lecturers, and students. The research uses the Rapid Application Development (RAD) method, which includes the stages of requirements planning, system design, prototype development, and testing. With ClassTIcs, staff can manage attendance data and schedules, lecturers can conduct digital attendance and upload lecture materials, and students can monitor lecture schedules and discuss through text forums. The results show that this application can help overcome the obstacles that often arise in the attendance process and management of study program activities while providing convenience for all actors involved. Further development is recommended to add other features to support the various needs of study programs in the future.

Application of the Moving Average Method in Forecasting Beverage Product Sales at Teman Cerita Cafe

Jamil Azhari Matondang, Ulfah Indriani — Tue, 28 Oct 2025 00:00:00 +0000

In the current digital era, the rapid growth of the culinary industry, particularly in the beverage sector, has created intense competition among small and medium enterprises. Café Kopi Teman Cerita in Medan is one of the businesses affected by fluctuating beverage sales, often experiencing challenges in predicting demand and managing inventory effectively due to its manual data processing system. This study aims to develop a beverage sales forecasting model using the Moving Average method to assist the café in optimizing raw material procurement and improving operational efficiency. The research was conducted by collecting historical sales data for various beverage products and applying the Moving Average technique to predict future sales trends. The evaluation results show that the forecasting model achieved satisfactory accuracy, with a Mean Absolute Error (MAE) of 10.17, Root Mean Squared Error (RMSE) of 12.26, and Mean Absolute Percentage Error (MAPE) of 20.59. These findings indicate that the Moving Average method can effectively capture sales fluctuations and produce reliable predictions for short-term demand planning. The implementation of this forecasting system is expected to help the café in making data-driven decisions, minimizing inventory discrepancies, and improving customer satisfaction. Moreover, the study provides insights into how simple yet effective analytical tools can support digital transformation and competitiveness in small-scale beverage businesses.

Data Mining Using Multiple Linear Regression Method for Stock Prediction

Methewkasly Pratama Sembiring, Wiwi Verina — Fri, 31 Oct 2025 00:00:00 +0000

This study aims to apply data mining techniques using multiple linear regression to predict inventory at PT. Sumber Jaya Motor's website. In today's digital era, companies face challenges in managing inventory, which can impact operational efficiency and customer satisfaction. Therefore, accurate inventory prediction is essential to improve inventory management efficiency. The multiple linear regression method was chosen because of its ability to link multiple independent variables with the dependent variable, thus providing more accurate predictions regarding required inventory. The data used in this study includes information related to sales, suppliers, and demand obtained from PT. Sumber Jaya Motor. The results of the multiple linear regression application indicate that the developed model can provide inventory predictions with a high degree of accuracy. This system is implemented on a website to facilitate real-time data-driven monitoring and decision-making. With the implementation of this method, it is hoped that PT. Sumber Jaya Motor can manage inventory more efficiently, reduce inventory costs, and improve customer service.

Application of AHP and Profile Matching Methods in Teacher Performance Assessment at XYZ School

Wahyu Hidayat, Ratih Puspasari — Sun, 26 Oct 2025 00:00:00 +0000

Teacher performance evaluation plays a critical role in improving educational quality; however, manual assessment systems often lead to subjectivity, inefficiency, and a lack of data-driven decision-making. This study aims to develop and implement a decision support system for teacher performance evaluation using the Analytical Hierarchy Process (AHP) and Profile Matching (PM) methods to produce more objective, transparent, and systematic assessments. The research was conducted at XYZ School with a sample of 20 teachers, evaluating five key performance aspects: learning planning, learning implementation, assessment and evaluation, classroom management, and communication with students. The AHP method was applied to determine the weight of each criterion through pairwise comparisons, while the Profile Matching method was used to align individual teacher competencies with ideal performance profiles. The system generated a ranking of teachers, identifying Drs. Hendarto Wijaya, Masriyanti, S.Pd, and Nurlia Syafina, S.Pd as the top three performers. The results indicate that combining AHP and PM effectively reduces subjectivity, enhances assessment accuracy, and accelerates the evaluation process. Furthermore, the web-based implementation allows automated reporting and easier data access, improving efficiency in teacher development planning. The implications of this study highlight that integrating multi-criteria decision-making models in educational management can strengthen evidence-based performance evaluation practices. Future studies should expand this framework across multiple institutions and incorporate advanced analytical methods to enhance system adaptability and scalability.

Determining Sales Based on Goods Data Classification Using the Web-Based C4.5 CRISP-DM Method

Bayu Gunawan, Wirhan Fahrozi — Fri, 31 Oct 2025 00:00:00 +0000

The increasing complexity of product distribution and sales activities at PT Mitra Bersama has created challenges in accurately classifying product performance, particularly due to manual data processing that is inefficient and prone to error. To address this issue, this study aims to develop an intelligent decision-support system capable of classifying best-selling and non-best-selling products using a data-driven approach. The CRISP-DM methodology was applied to guide the overall analytical process, consisting of business understanding, data understanding, data preparation, modeling, evaluation, and deployment. The C4.5 algorithm was used to perform the classification through entropy and information gain calculations to determine the most influential attributes. The results show that Type of Food has the highest information gain (0.0113340), followed by Initial Stock, Unit, Month, and Ending Stock, indicating that product characteristics and early inventory levels play a significant role in predicting sales performance. These findings were implemented into a web-based application to facilitate real-time classification and assist decision-makers in optimizing inventory planning, distribution strategies, and sales forecasting. This research contributes to improving organizational efficiency by providing a systematic, accurate, and accessible tool that supports better strategic decision-making in product sales management.

Design and Construction of Employee Attendance Using a Facial Recognition System at PT. Astra Daihatsu Krakatau

Ihsan Syahputra, Nita Syahputri — Tue, 28 Oct 2025 00:00:00 +0000

The rapid advancement of information technology has transformed organizational operations, including human resource management systems, where accuracy and efficiency in employee attendance recording are crucial for maintaining discipline and supporting performance evaluation. However, traditional attendance systems such as manual recording or fingerprint scanning remain limited, especially for employees who are working remotely or traveling, leading to data inconsistencies and delays. This research aims to design and implement a machine learning–based facial recognition attendance system to improve the flexibility, accuracy, and reliability of attendance processes at PT Astra Daihatsu Krakatau. The study employs the Unified Modeling Language (UML) approach for system design and utilizes facial recognition algorithms to automate attendance verification through biometric analysis. The resulting system comprises several key modules—login, main menu, employee data, attendance statistics, and history tracking—providing real-time and integrated attendance monitoring accessible from various devices. The findings indicate that the system effectively addresses the inefficiencies of conventional methods by enabling accurate biometric verification, minimizing fraud, and supporting remote attendance logging. This innovation enhances organizational transparency, operational efficiency, and adaptability in line with digital transformation initiatives. The implication of this study highlights the strategic role of artificial intelligence in modernizing workforce management and optimizing administrative processes within industrial environments.

Location Based Employee Attendance App with WFA Support and Real Time Location Monitoring

Muhammad Fahri Fahri, Elida Tuti Siregar — Fri, 31 Oct 2025 00:00:00 +0000

The attendance system is an important component in supporting the operational effectiveness of institutions, including the Mentari Meraki Asa Foundation (YMMA), which has a decentralized and dynamic work pattern. This study aims to develop a location-based employee attendance system with Work From Anywhere (WFA) features and real-time location monitoring. The system was designed using the Fishbone Diagram approach to analyze the root causes of problems and developed using web and Android-based technologies. The design process included identifying system requirements, implementing geofencing and GPS technologies for location validation, and establishing multi-layer authentication through additional verification for attendance outside the main location. Real-time location monitoring was implemented to enhance the accuracy of attendance data. The system was tested through functionality testing, location accuracy testing, and user experience testing. The results showed that the developed attendance system is capable of accurately recording attendance, supporting employee mobility, and improving operational efficiency. This research contributes to the development of modern attendance systems that are adaptive to flexible work patterns and can be implemented as a reference for future attendance system development.

Augmented Reality in Learning the Introduction to the Toba-Simalungun Batak Language Using the Marker-Based Tracking Method

Andreas Simbolon, M Irfan Aldy Nasution — Thu, 23 Oct 2025 00:00:00 +0000

Regional languages are invaluable cultural treasures and are inseparable from socio-cultural values that reflect the principles of regional behavior and identity. In the areas where tribes are spread in North Sumatra Province, there are several languages such as Batak Toba and Simalungun which are used in several areas in North Sumatra. Global developments have made the use of Batak Toba and Simalungun languages decrease and the intention to learn from children to adults is decreasing and the lack of learning facilities to understand Batak Toba and Simalungun languages. This study will aim to combine Batak Toba and Simalungun languages with augmented reality (AR) technology in order to obtain more efficient, interesting and interactive learning facilities for children to adults in learning Batak Toba and Simalungun languages. The application development method uses Marker Based Tracking as an augmented reality (AR) reference in the development of the Android-based Batak Toba-Simalungun Language Introduction Learning Application Marker Based Tracking Method. The Marker-Based Tracking method for the Toba-Simalungun Batak language, produced in this application, is a 3-dimensional (3D) image of the language. When the user points the camera at an object or 3D image, the application displays a 3-dimensional (3D) image of the Toba-Simalungun Batak language pattern captured by the user's camera.

Road Damage Level Prediction Using Data Mining-Based Regression Method

Rina Indah Widya Pangaribuan, Ratih Puspsari — Fri, 31 Oct 2025 00:00:00 +0000

Road damage is a critical infrastructure issue that significantly affects transportation safety, mobility efficiency, and vehicle operating costs, creating the need for an accurate and reliable prediction system to support timely maintenance planning. This study aims to develop and evaluate a road damage prediction model using a data mining–based linear regression method implemented in the WEKA environment. The dataset consists of 140 entries and includes key predictor variables such as daily vehicle volume, road age, rainfall, pavement type, and drainage quality. The methodology involves data preprocessing, linear regression modeling, and performance evaluation using metrics such as MAE, RMSE, and the correlation coefficient. The results show that the linear regression model demonstrates strong predictive capability, achieving a correlation coefficient of 0.8593, an MAE of 6.7954, and an RMSE of 7.763, with road age and pavement type identified as the most influential predictors. These findings indicate that linear regression is an effective and interpretable approach for modeling road deterioration levels and can be utilized as a decision-support tool in road maintenance planning and infrastructure asset management. Practically, the model provides data-driven insights for local governments and related agencies in optimizing repair scheduling and budget allocation based on predicted damage levels.

Product Demand Prediction in E-Commerce Systems Using the Monte Carlo Method

Fri, 31 Oct 2025 00:00:00 +0000

The rapid growth of e-commerce has intensified demand uncertainty, creating significant challenges in inventory management due to the risks of overstock and stockout conditions. Fluctuating consumer behavior and dynamic digital market trends require forecasting approaches capable of modeling probabilistic variability rather than relying solely on deterministic estimates. This study aims to analyze and implement the Monte Carlo simulation method to forecast product demand in an e-commerce system and to evaluate its effectiveness in supporting optimal inventory decision-making. The research adopts a quantitative approach using historical monthly sales data of laptop products collected over a ten-month period. The Monte Carlo method was applied by constructing probability distributions, calculating cumulative probabilities, defining random number intervals, and performing repeated simulations to generate demand predictions. The simulation results produced an average predicted demand of 139 units, closely aligned with the historical average of 137 units, with a Mean Absolute Deviation (MAD) of 2 units, indicating a low prediction error. These findings demonstrate that the Monte Carlo approach effectively captures demand variability and provides accurate probabilistic estimates. The study implies that integrating Monte Carlo simulation into e-commerce inventory planning can enhance risk-based decision-making, improve stock control accuracy, and reduce potential financial losses associated with demand uncertainty.

Comparison of the Waspas Method with the OCRA Method in Determining Web-Based Aid Recipients

Muhammad Randy Randy, Silvia Lestari — Tue, 28 Oct 2025 00:00:00 +0000

The determination of scholarship aid recipients at Pangeran Antasari High School has long relied on a manual and subjective process, leading to inefficiencies, inaccuracies, and potential biases in decision-making. To address this issue, this research aims to develop and compare the effectiveness of two multi-criteria decision-making (MCDM) methods—the Weighted Aggregated Sum Product Assessment (WASPAS) and the Operational Competitiveness Rating Analysis (OCRA)—within a web-based Decision Support System (DSS) for determining scholarship recipients. The study applies both methods using nine evaluation criteria, including socioeconomic and academic factors, to rank eligible students objectively. The results reveal that the WASPAS method produces more consistent, stable, and transparent outcomes, with scores that decline gradually and proportionally across alternatives, while the OCRA method demonstrates higher sensitivity to minor data variations, resulting in less stable rankings. Consequently, WASPAS proves to be more suitable for decision contexts that prioritize fairness, stability, and comprehensive evaluation of multiple criteria. The implementation of this method in an automated DSS enhances the objectivity, efficiency, and accountability of scholarship distribution processes. The study’s findings contribute to the advancement of decision-support frameworks in educational institutions and provide a methodological reference for future applications of MCDM models in resource allocation and policy decision-making.

Combination of Gada and GRA Methods in Decision Support System for Determining Permanent Employees

Riski Pransiskus Matondang, Charles Jhony Mantho Sianturi — Fri, 31 Oct 2025 00:00:00 +0000

Determining permanent employees is an essential aspect of human resource management, as it affects the stability, productivity, and sustainability of a company. This decision-making process often faces challenges such as subjectivity in assessment and limitations in processing employee performance data. Therefore, a Decision Support System (DSS) is needed to assist management in determining permanent employees in a more objective, accurate, and structured manner. This study proposes a combination of Gada and Grey Relational Analysis (GRA) methods in the design of a web-based DSS at PT. Medan Bajaindo. The Gada method is applied to determine the weight of criteria based on their importance level, while the GRA method is used for ranking employee alternatives by calculating their closeness to the ideal solution. The system is developed on a web-based platform, making it accessible for management. The test results show that the combination of Gada and GRA methods can provide more accurate recommendations for determining permanent employees compared to manual evaluation. Thus, this system is expected to support companies in making decisions that are more objective, transparent, and efficient.

K-Nearest Neighbor (K-NN) Method for Disturbance Classification of Customer Wifi Networks at PT. Global Karya Wanda

Muhairoh Indah Cahyani, Khairul Ummi — Thu, 23 Oct 2025 00:00:00 +0000

This study aims to develop and implement a Wi-Fi network disturbance classification system using the K-Nearest Neighbor (K-NN) algorithm at PT. Global Karya Wanda. The purpose of this research is to identify and classify Wi-Fi network conditions based on standard categories such as interference, troubleshooting, disconnection, and signal loss, thereby improving the efficiency and accuracy of network monitoring. The system was designed and developed using PHP and MySQL, with datasets obtained from PT. Global Karya Wanda’s operational network records. The classification process employed the K-NN algorithm to distinguish between Disturbance and Not Disturbance network states. The experimental results demonstrate that the K-NN method provides fast, automatic, and accurate classification performance, supporting the company in optimizing its troubleshooting workflow and enhancing customer service reliability. From a practical standpoint, the model enables more systematic network performance monitoring and proactive disturbance management. Scientifically, this research contributes to the application of machine learning algorithms in network performance analysis and telecommunications service optimization. Future studies are recommended to integrate hybrid approaches such as KNN–SVM or machine learning API integration to improve classification accuracy, scalability, and real-time responsiveness in larger and more dynamic network environments.

Developing Digital Examination System Security through the Implementation of the RC4 Algorithm

Muhammad Luthfi Moratuah Lubis, Iwan Fitrianto Rahmad — Fri, 31 Oct 2025 00:00:00 +0000

The increasing shift toward digital-based academic administration presents significant challenges for educational institutions, particularly regarding the security of examination data, which remains vulnerable when managed through manual processes. To address this issue, this study aims to develop and implement a secure web-based digital examination system using the RC4 cryptographic algorithm to ensure the confidentiality, integrity, and availability of assessment records at Harapan Mekar 1 Vocational School. Employing the CRISP-DM methodology, the research includes business understanding, data preparation, system development, encryption–decryption integration, and system validation. The RC4 algorithm, known for its lightweight structure and fast computational performance, was applied to encrypt student, teacher, and examination data across all operational modules. The results show that the system successfully encrypts and decrypts data through stable implementation of the Key Scheduling Algorithm (KSA) and Pseudo Random Generation Algorithm (PRGA), thereby strengthening protection against unauthorized access and eliminating risks associated with traditional manual data handling. The system effectively enhances operational efficiency and data security, demonstrating that RC4 remains a viable encryption option for educational environments with limited computational resources. This study offers practical implications for schools seeking accessible security solutions and serves as a reference for further system enhancements using more advanced cryptographic algorithms or extended digital examination features.

Decision Support System in Marketing Strategy Using Data Mining Techniques

Leliana Harahap, Sartika Dewi Purba, Jonas Franky R Panggabean, Kamson Sirait — Fri, 31 Oct 2025 00:00:00 +0000

The increasing complexity of market competition and the rapid growth of enterprise data have made traditional marketing decision-making approaches inadequate in addressing information asymmetry and dynamic market changes. Conventional decision support systems (DSS) are often limited to data-level reporting and lack advanced analytical capabilities to uncover hidden patterns and strategic insights. This study aims to design and evaluate an intelligent marketing Decision Support System by integrating data warehousing, Online Analytical Processing (OLAP), and data mining techniques to enhance the quality and effectiveness of marketing decisions. The proposed method adopts a data-driven DSS architecture that performs extract–transform–load (ETL) processes to build a unified subject-oriented data warehouse, followed by multidimensional analysis and knowledge discovery using decision tree classification and neural network models. Experimental validation was conducted using FoodMart 2000 sales data to assess the predictive performance and decision support capability of the system. The results demonstrate that the three-layer BP neural network model achieved a mean absolute percentage error of 15.13% in sales prediction, indicating satisfactory forecasting accuracy, while simulation and sensitivity analyses reveal a positive correlation between promotional investment and corporate profit growth. These findings confirm that the proposed marketing DSS can effectively reduce information asymmetry, improve forecasting reliability, and support strategic marketing decisions related to pricing, promotion, and market expansion. The study implies that integrating data mining with DSS provides a robust analytical foundation for data-driven marketing management and sustainable enterprise competitiveness in complex market environments.

Implementation of the Apriori Method in Consumer Purchasing Patterns at PT. Ouzen Anugerah Indonesia

Muhammad Alif Yustio, Ratih Puspasari — Sun, 26 Oct 2025 00:00:00 +0000

The rapid growth of the cosmetics industry has created intense competition among retail companies, including PT. Ouzen Anugerah Indonesia, which faces challenges in understanding consumer purchasing behavior due to the lack of systematic data analysis. This study aims to analyze consumer purchasing patterns and identify frequently purchased product combinations using the Apriori algorithm as a data mining technique. The research employed a descriptive quantitative approach based on 5,000 sales transactions recorded between January and December 2024, consisting of transaction IDs, dates, and purchased product lists. Through the Apriori algorithm, the study generated association rules that reveal relationships among products, represented by support and confidence values. The results show that several cosmetic items exhibit strong associative relationships, indicating consistent consumer purchasing tendencies that can be utilized for cross-selling strategies, promotional bundling, and inventory optimization. These findings highlight the effectiveness of data mining in transforming raw transaction data into actionable insights that support data-driven business decision-making. The study contributes theoretically by reinforcing the application of association rule learning in medium-scale retail contexts and practically by providing a framework for developing marketing and operational strategies based on data analysis. The research also suggests that future studies integrate broader datasets and additional variables, such as consumer demographics and temporal factors, to enhance analytical depth and model generalization.

Designing an Online Train Ticket Booking Website Using the Design Thinking Method

Dea Cania, Budi Triandi — Fri, 31 Oct 2025 00:00:00 +0000

The increasing demand for train services in Indonesia has highlighted significant challenges in conventional ticket reservation systems, including long queues, limited customer time, scheduling errors, and rapid ticket depletion during peak periods. This study aims to develop and evaluate a digital train ticket reservation application designed to streamline booking processes, improve user experience, and provide real-time access to ticket availability, pricing, and schedules. The research employs a user-centered Design Thinking approach to guide the development of key functional modules, including login, train and route management, schedule creation, order processing, and user profile management. Results demonstrate that the application successfully integrates these functionalities into a unified platform, reducing operational inefficiencies, minimizing user errors, and enabling users to book tickets conveniently from multiple devices. The system enhances administrative oversight through centralized order reporting and data management, while improving overall service reliability and user satisfaction. These findings imply that human-centered digital platforms can transform traditional public transportation services by increasing operational efficiency, scalability, and customer convenience. Furthermore, the study identifies areas for future improvement, such as incorporating predictive analytics for demand forecasting, dynamic seat allocation, and extended accessibility features to accommodate diverse user needs. The results provide practical insights for railway operators seeking to implement adaptive, reliable, and user-friendly ticket reservation systems, underscoring the value of integrating technology, usability, and real-time information in public transport services.

Design and Development of an Android-Based Inpatient Room Availability Application for Sinar Husni Hospital

Zuki Salwani D Salwani D, Lili Tanti — Mon, 20 Oct 2025 00:00:00 +0000

For every patient requiring inpatient care, Sinar Husni Hospital provides inpatient rooms in various classes. Sinar Husni Hospital also accepts medical and inpatient services through the Social Security Agency (BPJS). A problem is that inpatient rooms at Sinar Husni Hospital are often unavailable due to the large number of people requiring inpatient care and limited space. This is discovered after people arrive for treatment and the hospital informs them of available inpatient rooms, ultimately resulting in patients being referred to other nearby hospitals. Therefore, a method is needed for Sinar Husni Hospital to provide information to the public regarding inpatient room availability and for the public to receive information about the availability of inpatient rooms at Sinar Husni Hospital. The researcher's solution involves using computers and Android devices using the K-NN method to recommend inpatient rooms for patients. This research also aims to create an application that can manage inpatient room data and disseminate information about inpatient room availability to the public, allowing them to determine the availability of inpatient rooms. KNN can determine rooms that meet patient needs and desired facilities. The application consists of two users: the first, an administrator via a web server, who manages inpatient room data, and the second, a member of the public, who receives inpatient room availability information via an Android device.

Implementation of the IPA Method in Patient Satisfaction Services at the Dr. Komang Makes Indonesian Navy Hospital, Website-Based

Listyowati Purwopeni, Ulfah Indriani — Tue, 28 Oct 2025 00:00:00 +0000

Hospitals play a crucial role in providing high-quality healthcare services, particularly for inpatients who require intensive and continuous care. However, gaps often exist between patient expectations and perceived service performance, leading to dissatisfaction and inefficiencies in hospital operations. This study aims to evaluate the quality of inpatient services at Dr. Komang Makes Indonesian Navy Hospital using the Importance Performance Analysis (IPA) method to identify priority areas for improvement. The research employed both library and field studies, including observations, interviews, and questionnaires distributed to 100 inpatients. Data were analyzed by calculating performance and expectation scores across five service quality dimensions—reliability, assurance, responsiveness, empathy, and tangibility—to determine satisfaction levels. The findings revealed an average satisfaction index of 1.01, indicating that overall inpatient services are satisfactory, particularly in the assurance, empathy, and tangible dimensions. However, the reliability and responsiveness dimensions showed notable performance gaps, especially concerning doctor visit schedules, clarity of procedures, and response to patient complaints. These results highlight the need for targeted interventions, such as enhancing digital integration, improving communication systems, and increasing staff responsiveness. The implementation of continuous IPA-based evaluations is recommended to support systematic quality improvement and strengthen patient trust. This study contributes to the development of a patient-centered, technology-driven service model that enhances healthcare quality and organizational performance at Dr. Komang Makes Indonesian Navy Hospital.

Implementation of the CLCG Method in the Implementation of Final Semester Exams

Fahri Ramadhan, Helmi Kurniawan — Fri, 31 Oct 2025 00:00:00 +0000

The final semester examination is one of the essential instruments in measuring students’ competency achievement in schools. However, conventional exam implementation often faces challenges such as high operational costs, time-consuming correction processes, and the potential for cheating due to identical questions among students. This study aims to design and develop an Android-based examination application by implementing the Couple Linear Congruential Generator (CLCG) method as a question randomization mechanism. CLCG was chosen because it can generate random numbers with a more uniform distribution and longer period compared to the conventional Linear Congruential Generator (LCG), ensuring that each student receives a different set of questions while maintaining the same level of difficulty. The research was conducted at SMK Harapan Mekar 1 Medan using a prototype-based system development approach. The results show that the developed Android-based exam application runs effectively, successfully randomizes questions using the CLCG method, and facilitates teachers in managing the question bank and accelerating the grading process. Therefore, the application of the CLCG method in an Android-based digital exam system proves to improve efficiency, objectivity, and security in the implementation of final semester examinations in schools.

Comparison of K-Means Clustering and Fuzzy Tsukamoto Algorithms for Grouping Student Data of NU Medan Middle School Based on Academic Achievement

Abdul Rasyid Pahlevi, Amrullah Amrullah — Sun, 26 Oct 2025 00:00:00 +0000

The increasing need for data-driven decision-making in education has encouraged the use of intelligent algorithms to evaluate and classify student academic performance more effectively. However, differences in algorithmic approaches often lead to variations in interpretation and categorization outcomes. This study aims to compare the performance of the K-Means and Fuzzy Tsukamoto algorithms in clustering student achievement data at SMP NU Medan to determine which method provides a more accurate and interpretable classification model. The research employs quantitative analysis using students’ semester grades processed through Python and Microsoft Excel, where K-Means utilizes centroid-based clustering (75, 85, and 95) and Fuzzy Tsukamoto applies fuzzy logic with weighted membership values (0, 5, and 10). The results reveal that K-Means produces a more proportional and stable clustering structure, effectively differentiating student achievement levels within the same population, while Fuzzy Tsukamoto offers a simpler, rule-based classification system aligned with fixed academic standards. The findings indicate that K-Means is more suitable for analyzing relative performance variations, whereas Fuzzy Tsukamoto is better suited for absolute classification and administrative evaluation. Both methods are easily implemented and can be integrated into educational management systems to enhance instructional decision-making. The study implies that a hybrid combination of these two algorithms may provide a more comprehensive analytical framework for evaluating student performance.

Multiple Linear Regression Method in Product Stock Prediction at PT. Kartika Mandiri Abadi

Jhea Armaya, Muhammad Hari Ramadhan — Fri, 31 Oct 2025 00:00:00 +0000

Inaccurate inventory management often leads to stock shortages or surpluses, which impact operational efficiency and customer satisfaction in medium-sized distribution companies such as PT. Kartika Mandiri Abadi. This study aims to develop a website-based stock prediction system using the Multiple Linear Regression (MLR) method to produce more accurate stock estimates and support managerial decision-making. The research methods included collecting historical sales and product inventory data, designing a web-based system using the Unified Modeling Language (UML) model, implementing MLR to predict inventory levels based on independent variables such as monthly sales and initial inventory, and testing the functionality and accuracy of the system. The results show that the MLR-based inventory prediction system is capable of producing more stable and consistent estimates compared to manual methods, reducing the risk of stock shortages or surpluses, and facilitating management in inventory planning and distribution scheduling. The implementation of a web-based system provides real-time access, data visualization, and structured reports that support faster and data-driven decision making. These findings emphasize the importance of integrating statistical methods with information technology to improve operational efficiency and inventory planning in distribution companies. This research also opens up opportunities for further development by incorporating external variables or hybrid approaches to improve prediction accuracy in dynamic market conditions.

Implementation of the Least Square Method in Website-Based Product Sales Prediction

Alfa Rizki Alwi, Charles Bronson Harahap — Tue, 28 Oct 2025 00:00:00 +0000

Sales play a crucial role in determining the continuity and growth of a company. PT. Medan Bajaindo, which operates in the steel product sales sector, requires a system capable of accurately predicting future sales to support decision-making and inventory planning. Currently, sales data analysis is still conducted manually, which is inefficient and prone to errors in forecasting. This research aims to design and develop a web-based sales prediction system by applying the Least Square method. The Least Square method is chosen because it can predict data trends based on historical sales patterns with relatively low error rates. The system is developed using web-based programming languages and supported by a database for storing sales data. The testing results show that the developed system can automatically calculate predictions and present sales trend information more quickly, accurately, and structurally compared to the manual method. With this system, the company’s management can more easily plan sales strategies and stock procurement in a timely manner.

Mekro Bisnis Center IoT-Based Housing Security System with Automatic Barrier Using e-KTP Verification and Sound Indicators

Muhammad Aidil Dzikri, Maulia Rahman — Tue, 02 Dec 2025 00:00:00 +0000

Residential security is an essential aspect of ensuring the comfort and safety of residents. However, the use of conventional systems such as manual security guards and barrier gates without identity verification is still commonly found, including at the Mekro Business Center Housing Complex. Problems arising from these systems include inefficiency, the potential for unauthorized access, and limitations in manual surveillance by security personnel. To address these issues, an automated system is needed to manage residential access in real-time and enhance overall security. This study aims to develop a security system based on the Internet of Things (IoT) using an automatic barrier gate equipped with e-KTP verification via RFID and a sound indicator for real-time notifications. The system was designed using the Fishbone Diagram approach as a reference in designing the components and workflow of the security system. The tools used include an ESP32 microcontroller, RFID RC522 module, infrared sensor, buzzer, and DFPlayer Mini speaker, all integrated with the Blynk application and a monitoring website. The testing process was conducted experimentally using the blackbox method through a miniature physical prototype simulation in six different scenarios. Out of 35 total tests, the system responded correctly 30 times under online conditions, achieving a 100% success rate in scenarios involving authorized access, unauthorized access, intrusion attempts, emergency access, and control via the application. All tests under offline conditions failed due to the lack of internet connectivity. The results show that this system can improve the efficiency and reliability of automatic access control in residential environments and support the performance of security personnel.

System Application Design Inventory Management in Sales Using Genetic Algorithms

Reinhard Parningotan Siahaan, Budi Triandi — Sun, 26 Oct 2025 00:00:00 +0000

PT. Kao Indonesia, a company engaged in the production and distribution of consumer goods, requires an efficient inventory management system to ensure a smooth and responsive sales process. One of the main challenges faced is the discrepancy between stock levels and market demand, which often leads to overstocking or stockouts and ultimately financial losses. This study aims to design and develop an inventory management system application that optimizes stock levels using a Genetic Algorithm (GA). The GA method is employed to determine the optimal inventory quantity by analyzing historical sales data and evaluating various stock-level scenarios to find the most efficient solution. The application was developed using the PHP programming language and a MySQL database. A case study at PT. Kao Indonesia involving sales and product inventory data over a specific period demonstrates that the system effectively enhances stock management efficiency, minimizes inventory discrepancies, and supports more accurate and data-driven decision-making in the company’s inventory management process.

Implementation of the AHP Method in Decision Making for Stunting Management at the Sicanang Community Health Center UPT Using a Website

Robetauli Febriyanti Limbong, M. Irfan Aldy Nasution — Fri, 31 Oct 2025 00:00:00 +0000

Stunting remains a persistent public health challenge in Sicanang Village, where socioeconomic disparities and limited access to essential services continue to hinder effective intervention strategies. This study aims to identify and prioritize the key determinants of stunting using the Analytical Hierarchy Process (AHP) to support evidence-based decision-making for targeted interventions. Five criteria were evaluated—stunting prevalence, poor family conditions, maternal education, access to sanitation, and program support—based on expert judgment and field data. The results show that stunting prevalence (31.68%) and poor family conditions (29.43%) are the most influential factors, followed by program support (17.94%), maternal education (10.94%), and sanitation access (9.99%). The Consistency Ratio (0.0107) indicates that the weighting process is methodologically valid. At the sub-district level, Bahari Village and Pulau Bahagia obtained the highest priority scores (3.0161), while Pulau Sicanang ranked lowest (2.6776), highlighting the need for differentiated intervention strategies across locations. These findings underscore the importance of data-driven prioritization to optimize resource allocation, strengthen local health programs, and address the multidimensional determinants of stunting. The study provides a practical framework for policymakers to design targeted, context-specific interventions that can enhance the effectiveness of stunting reduction efforts in vulnerable communities.

Lightweight modular transporter box: structural validation and flight-range assessment

Ryan Rifat, AH Fauzi, NH Silalahi — Fri, 31 Oct 2025 00:00:00 +0000

Unmanned aerial vehicles (UAVs) are increasingly applied in logistics, yet their payload and endurance remain key constraints. This research addresses the problem by developing a lightweight modular transporter box for the DJI Matrice 600 hexacopter. The objective was to ensure structural reliability while maintaining operational flight range. A quantitative approach was applied: design requirements and objectives (DR&O) were established, five 3D design iterations were created, and finite element analysis (FEA) was performed to evaluate stress, deformation, and safety factors. Polyethylene terephthalate (PET) was selected as the construction material due to its strength-to-weight ratio and recyclability. Endurance analysis incorporated empirical flight speed and battery degradation models to assess the effect of payload integration. The final prototype achieved a mass of 1.79 kg, carried a 5 kg payload, and maintained a factor of safety of 4.46. Flight analysis confirmed an operational range of 18.8 km in 16.8 minutes, exceeding the minimum 11 km requirement. These results demonstrate that PET-based transporter boxes can provide lightweight, modular, and structurally safe payload solutions without significantly reducing UAV endurance. The study contributes to UAV logistics applications in defense and emergency delivery, and future work will involve prototype fabrication and experimental validation.

Application of Book Clustering Based on Borrowing Frequency Using K-Means at PBD Aviation Vocational School

Nurul Nazli, Linda Wahyuni — Tue, 28 Oct 2025 00:00:00 +0000

The rapid advancement of science and technology has enabled the application of data mining techniques in various domains, including education and library management. At PBD Aviation Vocational School, managing the library collection presents a major challenge due to the large number of students and the diverse frequency of book borrowing, leading to inefficiencies in procurement and maintenance. This study aims to develop a data-driven system for clustering books based on borrowing frequency using the K-Means algorithm to optimize collection management. The research was conducted using a quantitative approach, where library borrowing data were processed through the K-Means clustering method integrated into a web-based system. The algorithm classified books into three categories: frequently borrowed, quite frequently borrowed, and rarely borrowed. The results showed that the implementation of the K-Means algorithm effectively identified borrowing patterns, enabling the library to make more accurate decisions regarding book procurement, maintenance, and collection renewal. Furthermore, the web-based interface facilitated faster access and real-time visualization of borrowing trends, improving operational efficiency and data-driven decision-making. The findings highlight the importance of integrating data mining methods in educational library systems to enhance resource utilization and service quality, supporting evidence-based management in the digital transformation era.

New Employee Recruitment Using Web-Based Oreste Method

Monica Nababan Nababan, Nurhayati Nurhayati — Fri, 31 Oct 2025 00:00:00 +0000

Decisions taken in the long term will affect the quality and performance of the company in the long term. PT. Intan Havea Industry, as a rapidly growing company, faces challenges in selecting candidates who best meet the required criteria. The employee selection process is still subjective and time-consuming, so a system is needed that can support decision making objectively and efficiently. This study aims to design and develop a Decision Support System (DSS) that can help PT. Intan Havea Industry in selecting new employees more effectively and efficiently. The method used in this system is ORESTE (Organization, Ranking, and Evaluation of Socio-Technical and Environmental Factors), which is able to evaluate various relevant criteria in the employee requirements process. The ORESTE method was chosen because of its ability to handle multi-criteria problems and provide objective results based on comparisons between candidates. The system is built web-based, allowing flexible access and can be integrated with various devices. This system accepts input data in the form of assessment criteria for each candidate and interview results, and provides candidate rankings sorted based on the evaluation value of the ORESTE method. System testing shows that this application can provide recommendations for the right candidates based on objective analysis, minimize subjective bias, and speed up the selection process. The results of this study are expected to contribute to the development of a web-based information system that can support more accurate decision making in the employee recruitment process at PT. Intan Havea Industry and other companies. With this system, it is hoped that the employee recruitment proces

Application for Searching for Livestock Distributor Locations at the North Sumatra Plantation and Livestock Service Using the Euclidean Distance Method

Prayogi Aulia, Elvin Syahrin — Thu, 23 Oct 2025 00:00:00 +0000

The Plantation and Livestock Agency of North Sumatra Province plays an important role in supporting the availability and distribution of livestock for farmers and livestock business actors. However, information about the locations of livestock distributors, which are scattered across various regions, is often difficult to access quickly and accurately. This situation makes the process of finding distributor locations inefficient and time-consuming. Therefore, a system is needed to help users find the nearest livestock distributor locations automatically and accurately. This study aims to design and develop an Android-based application that provides information on livestock distributor locations in North Sumatra Province and determines the nearest distance between the user and the distributors using the Euclidean Distance method. This method is used to calculate the distance between two points based on their geographic coordinates (latitude and longitude), allowing users to get recommendations for the nearest distributor from their current position. The result of this study is an Android application that can display a list of livestock distributors, show their locations on a map using GPS services, and provide real-time information about the closest distance. This application is expected to make the process of finding livestock distributor locations more effective and efficient, as well as help the Plantation and Livestock Agency of North Sumatra Province provide digital information services to the public.

Implementation of the TPM Method to Increase Production Machine Productivity at PT. Cemara Saringan Medan

Elisa Lbn Raja, Asbon Hendra Azhar — Fri, 31 Oct 2025 00:00:00 +0000

PT. Cemara Saringan Medan, a medium-scale manufacturer of palm oil filtration components, continues to experience frequent machine downtime and performance inefficiencies that disrupt production stability and increase operational costs. To address these challenges, this study aims to enhance machine reliability and productivity through the implementation of Total Productive Maintenance (TPM) integrated with a web-based monitoring system. Using a waterfall-based research methodology, the study involved data collection, preprocessing, TPM computation (availability, performance efficiency, and quality rate), OEE analysis, UML-based system design, application development, and black-box testing. The results show that the company’s production machines achieved an availability level of 89%, performance efficiency of 83%, and a product quality rate of 96%, indicating that productivity losses are primarily attributed to breakdown-related downtime and cycle-time inefficiencies rather than product defects. The integration of TPM with digital monitoring further demonstrates strong potential to enhance maintenance transparency, reduce human error, and support more responsive decision-making. These findings imply that adopting web-based TPM tools can significantly improve operational control and serve as a scalable foundation for predictive maintenance in medium-scale manufacturing environments.

Evaluation of the usability of the faculty performance assessment questionnaire using the system usability scale (SUS) method

Sutrisno Situmorang, Jontinus Manullang, Harmoko Lubis, Aditiarno Manik — Fri, 31 Oct 2025 00:00:00 +0000

The digital transformation of lecturer performance evaluation systems in higher education has improved efficiency in data collection and reporting; however, system effectiveness depends significantly on its usability. A system that is difficult to use may reduce student participation and compromise the quality of evaluation data. This study aims to evaluate the usability level of an online lecturer performance assessment questionnaire system using the System Usability Scale (SUS). The research employed a descriptive quantitative approach involving 30 active students as respondents who had completed the online questionnaire. Data were collected using a 10-item SUS instrument with a five-point Likert scale and analyzed according to standard SUS scoring procedures, followed by descriptive interpretation and validity testing using Pearson correlation. The results showed that the average SUS score was 49.75, which falls into the “Poor” category, indicating that the system’s usability level has not yet reached an acceptable standard. Although the system is generally accessible and relatively easy to learn, several aspects—particularly navigation clarity, interface consistency, and user feedback mechanisms—require improvement. The validity test confirmed that all questionnaire items were statistically valid. These findings imply that systematic redesign and iterative usability evaluation are necessary to enhance user experience, increase student participation, and strengthen the effectiveness of lecturer performance evaluation as part of sustainable academic quality assurance in higher education institutions.

Diet Recommendation System for Kidney Disease Patients Using Collaborative Filtering

Bintang Cimdy Lorenza Sitompul, Ratih Puspasari — Tue, 28 Oct 2025 00:00:00 +0000

Chronic kidney disease (CKD) remains a major global health challenge, particularly in Indonesia, where limited awareness and inadequate dietary management contribute to the progression of renal complications. Patients often face difficulties in selecting foods that meet both medical and nutritional requirements, underscoring the need for intelligent dietary guidance. This study aims to develop a personalized dietary recommendation system for kidney failure patients using a hybrid approach that combines content-based and collaborative filtering techniques. The model was designed to analyze patients’ food preferences, nutritional composition, and health conditions to generate appropriate dietary recommendations. The system’s performance was evaluated using cosine similarity and predictive accuracy metrics, including RMSE, precision, recall, and F1-score. The results show that the proposed model achieved an accuracy of 83%, precision of 75%, recall of 100%, and F1-score of 86%, demonstrating its effectiveness in identifying dietary similarities and preferences among patients with comparable clinical profiles. Furthermore, by integrating nutritional content data such as sodium, potassium, and protein levels, the system successfully provided clinically safe and personalized recommendations aligned with renal dietary guidelines. These findings highlight the potential of artificial intelligence–based recommendation systems to support dietitians in improving the accuracy and efficiency of nutritional counseling, thereby promoting patient adherence and enhancing the quality of kidney disease management in hospital settings.

Application of the Fuzzy Time Series (FTS) Method to Estimate the Increase in Stunting at the North Sumatra BKKBN

Elisabet Diadora Manurung, Ahir Yugo Nugroho Harahap — Fri, 31 Oct 2025 00:00:00 +0000

Stunting is one of the chronic nutritional problems that remains a major concern in Indonesia, particularly in North Sumatra Province. This condition occurs due to prolonged nutritional deficiencies that affect children’s physical growth and cognitive development. The National Population and Family Planning Agency (BKKBN) of North Sumatra plays an essential role in monitoring and reducing stunting rates. However, the prediction methods currently used are still conventional and have not utilized modern predictive technologies. As a result, the forecasting accuracy is often low because stunting data tends to be fluctuating and influenced by various dynamic factors. This study aims to apply the Fuzzy Time Series (FTS) method to predict the increase in stunting cases at BKKBN North Sumatra. The FTS method is selected because it can handle data uncertainty and non-linear patterns, making it more effective in recognizing trends in stunting cases over time. The data used in this study were obtained from BKKBN North Sumatra, consisting of the number of stunting cases in 2023 and 2024 across 33 districts and cities. The research stages include data collection, fuzzification, fuzzy relation formation, defuzzification, and the implementation of the results into a web-based application using PHP programming and MySQL as the database. The results show that the Fuzzy Time Series method can effectively predict the upward and downward trends of stunting cases, where 16 districts/cities experienced an increase and 17 districts/cities experienced a decrease in stunting numbers for the following year. The developed system helps BKKBN analyze data more quickly, accurately, and efficiently. Therefore, the application of the Fuzzy Time Series method proves to be effective as a decision-support tool for planning prevention and mitigation programs for stunting in North Sumatra.

Executive Information System for the Prevention and Handling of Sexual Violence in School Environments

Nur Falah Akbar Nst, Muhammad Fauzi — Mon, 20 Oct 2025 00:00:00 +0000

Sexual violence in schools is a serious issue that can negatively impact the mental and physical well-being of students. The lack of an integrated system for reporting and monitoring cases often results in suboptimal prevention and response efforts. This study aims to design and implement an executive information system that assists school management, especially decision-makers, in making data-driven decisions regarding the prevention and handling of sexual violence. The system development follows the waterfall method, consisting of requirements analysis, system design, implementation, and testing phases. The result of this research is a web-based system with an interactive dashboard that displays case reports, incident trends, and follow-up recommendations. The system was tested using usability testing and the results indicated that the system is user-friendly and effectively supports the decision-making process. The implementation of this executive information system is expected to enhance responsiveness, transparency, and effectiveness in addressing sexual violence within the school environment.

Application of the C4.5 Algorithm Method to Identify Factors Affecting Package Delivery Delays at the Post Office

Muhammad Taufik Siahaan, Rahmad Doni — Thu, 30 Oct 2025 00:00:00 +0000

In the modern logistics landscape, timely package delivery has become a critical determinant of service quality and customer satisfaction. However, frequent delivery delays at the Post Office continue to pose operational challenges, largely due to multifactorial causes such as weather, distance, and scheduling inefficiencies. This study aims to identify and analyze the dominant factors influencing delivery delays using the C4.5 decision tree algorithm, a robust data mining method capable of handling categorical and continuous variables while generating interpretable decision rules. The research utilized historical delivery data from the Post Office, encompassing attributes such as weather conditions, delivery distance, order time, and package type. The analysis revealed that weather conditions had the highest information gain (0.0282430), indicating their dominant impact on delivery performance, followed by distance and package characteristics. The model successfully generated 112 decision rules that enable managers to predict and mitigate potential delays. The findings highlight the effectiveness of the C4.5 algorithm in uncovering complex patterns within operational data and its potential to support data-driven decision-making in logistics management. The implementation of this model can significantly enhance delivery reliability, operational efficiency, and customer trust, representing a strategic advancement toward digital transformation in postal services.

Application of the Trend Moment Method in Product Sales Forecasting at TB. Mitra Baru

Prans Tito, Asbon Hendra Azhar — Fri, 31 Oct 2025 00:00:00 +0000

Advances in information technology have made it easier for companies to manage sales data and conduct forecasting to support decision-making. TB. Mitra Baru, a trading company, often faces challenges in predicting future product sales, which impacts stock availability and business planning. This study aims to design and implement a website-based sales forecasting system using the Trend Moment method, a time series method that utilizes a trendline equation based on the average value over time and sales volume. Sales data is used as input to generate predictions for the next period in a systematic and structured manner. Test results indicate that the Trend Moment method is capable of providing sales estimates with a relatively low error rate, thus providing a basis for consideration in inventory management and sales strategies. With this website-based system, TB. Mitra Baru can easily access actual sales reports and forecast results at any time in real time, thus supporting operational effectiveness and more informed decision-making.