Jurnal Media Computer Science

Classification of Eligibility for Prospective Recipients of the KIP Kuliah Scholarship Using the K-Nearest Neighbor Method at Dehasen University in Bengkulu

Imam Ashari, Herlina Latipa Sari, Devi Sartika — Sun, 25 Oct 2026 15:44:10 +0700

The Kartu Indonesia Pintar (KIP) Kuliah program aims to help high-achieving students from low-income backgrounds continue their education at the university level. At Dehasen University in Bengkulu, the high number of applicants far exceeds the available quota, while the selection process is still conducted manually by reviewing applications one by one. This situation results in a slow and inefficient selection process that is prone to human error and potentially introduces subjectivity in determining the eligibility of scholarship candidates.This study aims to apply the K-Nearest Neighbor (KNN) algorithm to classify the eligibility of KIP Kuliah scholarship candidates based on socioeconomic and academic data. The research stages included importing a dataset of 665 records, data cleaning, data selection, data transformation, Min-Max normalization, and splitting the data into training and testing sets, followed by classification by testing values of K = 3, 5, 7, and 9 using Euclidean distance calculations. The research results show that the KNN algorithm was successfully applied in the classification process, with the best accuracy achieved at K=7: 63.91%, Precision 53.90%, Recall 63.91%, and F1-Score 57.61%. Thus, the K-Nearest Neighbor (KNN)-based classification system developed can facilitate the initial selection process for prospective recipients of the KIP Kuliah Scholarship more quickly, consistently, and objectively compared to a manual selection process.

Predicting Consumer Demand At Nayla Buah Store, Ratu Agung District, Bengkulu City Using Machine Learning Technology

Yulia Darnita, Muntahanah Muntahanah, Dandi Sunardi, Ahmad Fauzi — Sun, 25 Oct 2026 15:44:40 +0700

The retail industry, including fruit stores, is significantly influenced by fluctuating consumer demand dynamics. An inability to accurately predict demand can lead to various operational issues for fruit stores, namely overstocking and understocking. Therefore, the ability to accurately forecast demand is crucial for business owners. The Random Forest Regressor and K-Nearest Neighbor (KNN) methods provide an evaluation of which algorithm is most suitable for predicting demand. Research results indicate that factors such as fruit type, day of the week, promotions, and weather influence the volume of consumer demand for fruit. These variables can serve as a basis for predicting fruit demand at the store. The prediction results from both models can be used to inform stock planning and operational decision-making. By forecasting daily demand more accurately, the store can mitigate the risk of losses associated with both overstocking and understocking. Overall, the Random Forest Regressor is the most recommended model for predicting fruit demand at Nayla Buah Store, as it offers higher accuracy and stability, and is capable of processing data variations more effectively.

Implementation of the Greedy Algorithm for Subject Scheduling at SMP Negeri 57 North Bengkulu

Yessi Indah Radianti, Reno Supardi, Desi Mahdalena — Mon, 26 Oct 2026 21:20:18 +0700

Subject scheduling is one of the essential activities in supporting the teaching and learning process in schools. Manual scheduling often presents several challenges, such as requiring considerable time, difficulties in allocating class periods, and the possibility of schedule conflicts. This study aims to develop a web-based subject scheduling system by implementing the Greedy Algorithm at SMP Negeri 57of North Bengkulu. The system was developed using the PHP programming language and MySQL database. The research methods included requirements analysis, system design using Data Flow Diagrams (DFD) and Entity Relationship Diagrams (ERD), system implementation, and functionality testing using the Black Box Testing method. The Greedy Algorithm was implemented to assist the scheduling process by gradually selecting available scheduling alternatives until a feasible schedule was generated. The results showed that the system was capable of managing teacher, subject, class, and class-period data in an integrated manner and generating lesson schedules more quickly and systematically than the manual process. With the implementation of this system, the subject scheduling process at SMP Negeri 57of North Bengkulu became more effective and efficient.

An Application For Predicting Children’s Nutritional Status In The Free Nutritious Meals Programme Using Machine Learning

Muntahanah Muntahanah, A.R. Walad Mahfuzi, Pahrizal Pahrizal, Sinta Yopita Utari — Mon, 26 Oct 2026 21:27:25 +0700

The Free Nutritious Meal Program (Program Makan Bergizi Gratis - MBG) serves as a strategic initiative to enhance human resource quality from an early age through targeted nutritional intervention for school children. However, implementation efficiency at the school level remains hindered by conventional and manual monitoring of student nutritional status. This manual approach risks calculation errors in anthropometric indices and delays the early detection of nutritional issues. This study aims to design, develop, and evaluate the reliability of a web-based decision support system capable of automatically predicting elementary school students' nutritional status to support targeted distribution within the MBG program. The modeling technique utilizes the K-Nearest Neighbor (KNN) algorithm, processing four primary input variables: gender, age (months), weight (kg), and height (cm), in accordance with the Body Mass Index-for-Age (BMI-for-Age) anthropometric standards established by Permenkes RI No. 2 of 2020. The data preprocessing stage incorporates automatic unit conversion of height from centimeters to meters (Height_meters = Height_cm / 100), Label Encoding, and Min-Max Scaling to normalize attribute scales and prevent variable dominance during Euclidean Distance calculation. Computational performance testing was conducted using a 4 x 4 Multi-class Confusion Matrix with a K=5 configuration on test samples from SDN 27 Bengkulu City. Experimental evaluation results demonstrate that the system achieved an Overall Accuracy of 90.91%, a Macro Average Precision of 96.88%, and a Macro Average Recall of 87.50%. These metrics prove that the KNN modeling approach possesses high precision and reliability in predicting four nutritional categories: thinness, normal, overweight, and obesity. In conclusion, the developed decision support application is scientifically validated and suitable for deployment as a digital ecosystem for schools and program managers to monitor students' physical health dynamics accurately, consistently, and in real time.

Design And Development Of A Qr Code-Based Teacher Attendance System At State Primary School 50 Bengkulu City

Mon, 26 Oct 2026 21:27:49 +0700

The process of recording teacher attendance at SD Negeri 50 Kota Bengkulu still uses an attendance book, resulting in manual data management. This study aims to develop a web-based teacher attendance system using QR Code and location validation. Data were obtained through observation, interviews, and literature study. The system was developed using the Waterfall method with PHP and MySQL and Visual Studio Code. The admin can manage teacher data, schedules, QR Codes, attendance data, and reports, while teachers record their attendance by scanning the QR Code and logging in using a username and password. The system automatically displays the schedule and validates the attendance time and location based on a predetermined polygon area. The system also limits attendance recording to once per day and provides a 15-minute late tolerance. The implementation results show that the system can perform attendance recording, location and time validation, data storage, and report generation, which can be printed and exported to Excel. Testing using Black Box Testing shows that the main system functions operate according to the requirements. The developed system can assist the school in digitally recording and managing teacher attendance and provide data for administrative needs.

Web-Based Attendance Application Using The Geolocation Method At Talang Perapat Village Office, West Seluma District

Mon, 26 Oct 2026 21:28:08 +0700

The development of information technology has encouraged the implementation of digital attendance systems in various government institutions to improve the effectiveness of employee attendance data management. However, the attendance system used at Talang Perapat Village Office, West Seluma District, still has limitations in accurately validating employee attendance. This study aims to design and develop a web-based attendance application using the geolocation method at Talang Perapat Village Office, West Seluma District. The application is designed to support a more effective, accurate, and transparent attendance recording process through the utilization of geolocation technology and photo documentation as attendance validation. The methods used in this study include observation, interviews, and literature review, while the system development process applies the Waterfall method. The resulting system is capable of recording location-based attendance, capturing photos as proof of attendance, and monitoring attendance data in real time. The developed application has proven to enhance the validity of employee attendance data, prevent attendance fraud through location and photo verification, and accelerate the attendance report management process for administrators at the Talang Perapat Village Office, West Seluma District.

Classification Of Social Assistance Eligibility In Talang Benuang Village Using C4.5 And Naive Bayes

Felly Intania, Yupianti Yupianti, Eko Suryana — Mon, 26 Oct 2026 21:28:30 +0700

The selection of Non-Cash Food Assistance (BPNT) recipients in Talang Benuang Village is still performed manually, creating a risk of mistargeted distribution. This study develops and compares C4.5 and Naive Bayes models for classifying BPNT eligibility. A data-driven approach followed CRISP-DM, using 100 labeled public records for modeling and 50 village household records for application. The analysis comprised data cleaning, ordinal transformation, pipeline construction, C4.5 hyperparameter tuning with GridSearchCV, and 5-fold cross-validation evaluated through a confusion matrix and ROC-AUC. Both algorithms achieved 93.00% accuracy, 100.00% precision, 85.42% recall, and a 92.13% F1-score. C4.5 was selected as the final model because it produced interpretable decision rules and an AUC of 0.986, exceeding the Naive Bayes AUC of 0.868. Application to 50 households yielded 42 eligible and 8 ineligible predictions. These findings indicate that C4.5 can support objective preliminary screening of BPNT recipients; however, administrative decisions still require verification through village deliberation and official ground-truth records.

Implementation of Sorting and Binary Search Algorithms in a Point-of-Sale (POS) Application to Optimize the Search for Basic Food Product Data

Antonius Antonius, Liza Yulianti, Ricky Zulfiandry — Mon, 26 Oct 2026 21:28:46 +0700

This study aims to implement the Selection Sort and Binary Search algorithms in a mobile-based Point of Sales (POS) application to improve the speed and efficiency of grocery product retrieval at Toko Rifki Kampung Melayu. The study employed Research and Development (R&D) with a software engineering approach. Data were collected through observation, interviews, documentation, and literature review. Selection Sort was used to arrange products alphabetically, while Binary Search was applied to locate products in the sorted array. The system uses a local SQLite database and can operate offline. Black-Box Testing of seven scenarios achieved a 100% Success Rate. Performance testing showed that Selection Sort sorted 13 items with 100% accuracy in 1.2–2.5 ms. Binary Search located the target in three iterations with a search time of less than 1 ms, while the theoretical maximum for 13 items is four iterations. These results indicate that the combination of the two algorithms effectively supports fast and accurate product retrieval and improves transaction efficiency.

The Application of Logistic Regression to Public Service Satisfaction in Tangga Batu Village

Muhamad Aldilah Julianto, Lena Elfianty, Achmad Fikri Sallaby — Mon, 26 Oct 2026 21:29:02 +0700

From a government administration perspective, public services at the village level are often assessed in a conventional and subjective manner. The purpose of this study is to evaluate the effectiveness of using the Binary Logistic Regression method to model and classify the level of community satisfaction with public services in Tangga Batu Village, Seluma Regency. This study employs a quantitative approach, using primary data collected from 276 respondents via a structured SERVQUAL questionnaire based on five dimensions: Tangible Evidence (X1), Reliability (X2), Responsiveness (X3), Assurance (X4), and Empathy (X5), all of which were validated using Pearson’s correlation test. Modeling was performed using the Python programming language, with 80% of the data allocated for training and 20% for testing. The results showed that the Binary Logistic Regression model was able to classify community satisfaction with an accuracy of 67.9% on the test data and an AUC-ROC of 0.755, as well as an average accuracy of 65.2% in the 5-fold cross-validation test. Based on the Wald test, the variables Empathy (X5) and Assurance (X4) were found to have a statistically significant effect on citizen satisfaction (p<0.05), with Odds Ratios of 1.8787 and 1.6610, respectively, while the other three variables were not significant. With an average predicted response time of 0.9795 milliseconds per sample, this model proved to be highly computationally efficient. This study concludes that the binary logistic regression method can be utilized by the Tangga Batu Village Government as an objective, data-driven decision-making tool to determine priorities for service improvements, particularly in enhancing the friendliness and personal attention of staff, as well as the reliability and transparency of administrative services.

Design And Development Of An Android-Based Automated Book Lending System Using Finite State Automata (FSA) And A Rule-Based Algorithm At The Bengkulu City Library And Archives Office

Ronaldi Saputra, Dewi Suranti, Ila Yati Beti — Mon, 26 Oct 2026 21:29:20 +0700

Manual book circulation management at the Bengkulu City Library and Archives Office frequently results in stock discrepancies, complex record-keeping, and a high incidence of overdue book returns. This study aimed to develop an Android-based automated library application integrating QR Code scanning to improve the efficiency and accuracy of circulation operations. The system employed Deterministic Finite Automata (DFA) to control book and transaction status transitions deterministically, a Rule-Based Algorithm to validate lending business rules, and atomic Cloud Firestore transactions to ensure data consistency based on ACID principles. In addition, a Dual Notification Strategy was implemented, consisting of direct notifications to administrators and automated local reminders scheduled for two days (H-2) and one day (H-1) before the due date. Black Box testing demonstrated a 100% functional success rate, with an average transaction response time of less than 3 seconds per scan. The results indicate that the developed system successfully minimizes human error, eliminates the risk of duplicate transactions, and provides practical support for the accountable and sustainable digitalization of library services.

An Expert System For Diagnosing Dengue Fever, Malaria, And Chikungunya Using An Integrated AHP And Certainty Factor Method

Irfan Fauzi, Sapri Sapri, Jhoanne Fredricka — Mon, 26 Oct 2026 21:29:39 +0700

Dengue Hemorrhagic Fever (DHF), malaria, and chikungunya are diseases commonly found in tropical regions and share several similar symptoms, which can make early identification difficult. This study aimed to design and develop a web-based expert system capable of assisting users in the initial diagnosis of DHF, malaria, and chikungunya based on the symptoms experienced. The system was developed by integrating the Analytical Hierarchy Process (AHP) and Certainty Factor (CF) methods. The AHP method was used to determine the priority level of each symptom based on expert assessments, while the CF method was used to calculate the degree of confidence in the resulting diagnosis. The system was developed using PHP programming language and a MySQL database. System testing was conducted using Black Box Testing and accuracy testing. The results showed that all system functions operated according to user requirements, and the system achieved an accuracy rate of 100% based on 10 test cases compared with expert diagnoses. The findings indicate that the system is capable of providing preliminary diagnoses along with their confidence levels quickly and systematically, making it suitable as a supporting tool for disease consultation.

Classification of Facial Skin Problems Using Convolutional Neural Networks (CNN) and Content-Based Filtering for Skincare Product Recommendation

Mila Cahyani, Sefrika Entas, Faisal Faisal — Mon, 26 Oct 2026 21:29:55 +0700

Many consumers find it difficult to independently identify their facial skin problems due to limited access to dermatological services. This study aims to classify seven facial skin problems using a CNN with ResNet50V2 and generate skincare product recommendations using Content-Based Filtering with TF-IDF and Cosine Similarity. The dataset consists of 2,450 facial skin images, with 350 images for each class. The model was trained through three progressive fine-tuning phases using a Dual Pooling classification head and Test-Time Augmentation, complemented by input validation through face detection and a confidence-score threshold. The results show that the model achieved a test accuracy of 82.53% and a macro F1-score of 0.82. The acne class achieved the highest performance with an F1-score of 0.89, while the dull skin/dark spots class obtained the lowest F1-score of 0.78, which was associated with visual similarities to signs of aging. The recommendation system generated product recommendations with similarity scores ranging from 0.278 to 0.321 based on the relevance of product attributes and active ingredients. Black-box testing showed that all tested system scenarios were valid.

Design Of A Web-Based Point Of Sale System For A Photo Studio: A Case Study Of Ruang Rana Palembang

Anisa Julianti, Muhammad Ihsan, Lucky Indra Kesuma — Mon, 26 Oct 2026 21:30:19 +0700

Ruang Rana Palembang is a photostudio business that still records transactions, customer data, reports, and receipts by hand. This practice invites recording mistakes, duplicate data, and slow access to information. This study designs a web-based cashier system that brings transactions, customer data, service menus, promotions, studio rooms, and reports together in one application. Development uses the sequential Waterfall model, running from requirements analysis through design and implementation to testing, evaluation, and maintenance. Observation, interviews, and a review of existing documents supplied the requirements, which were then modelled in three Unified Modeling Language (UML) diagrams: use case, activity, and class. Two kinds of users, the cashier and the owner, work with the system under separate access rights. The result is a cashier system with login, session- and duration-based transactions, payment with automatic change calculation, receipt printing, customer and transaction data that can be printed as PDF or exported to Excel, management of menus, promotions, rooms, and users, and monthly revenue reports for the owner. Its data structure consists of eight classes, and functional testing is planned with Black Box Testing. The system is expected to shorten transaction time, make records more accurate, and give the owner a firmer basis for business decisions.

Design and Development of the Vocaup! Website for English–Indonesian Vocabulary Learning Based on Flashcards and Quizzes

Nabilah Dwi Agustin, Zulhipni Reno Saputra Elsi — Mon, 26 Oct 2026 21:30:41 +0700

Conventional English vocabulary learning methods, such as memorizing word lists, are often considered less effective because they offer little interactivity and provide no immediate feedback. This study aims to design and develop the Vocaup! website, an English–Indonesian vocabulary learning platform equipped with interactive flashcards, quizzes, pronunciation audio, and progress tracking. The system was developed using a Research and Development (R&D) approach with the Waterfall model, covering requirements analysis, system design using UML, implementation using native PHP and MySQL, black box testing, and evaluation and improvement. The system provides illustrated flashcards with audio, quizzes with translate, listening, and arrange-words question types, a guest mode, login to store learning progress, badges as a gamification element, and an admin panel to manage user data, vocabulary, and quiz questions. The results show that the Vocaup! website provides an interactive and easy-to-use vocabulary learning medium, and all features work according to user needs and support independent learning. The website is expected to serve as an accessible digital learning alternative for a wide range of learners.

Predicting Stunting Levels in Toddlers Using Machine Learning Based On Health Data and Image Processing

Gathan Raihan Al Saleh, Juju Jumadi, Ahmad Arib Al Farisy — Mon, 26 Oct 2026 23:50:21 +0700

Stunting is one of the most prevalent chronic nutritional disorders affecting toddlers, with significant impacts on their physical growth and cognitive development. The limited quality and completeness of health records collected manually at community health posts (Posyandu) present challenges for the early identification of stunting. This study aimed to develop a predictive model for stunting levels in toddlers using a machine learning approach based on health data and image processing. The theoretical framework incorporated Machine Learning, Support Vector Machine (SVM), Random Forest, Convolutional Neural Network (CNN), Image Processing, and Ensemble Learning using the Hard Voting technique. This study employed a quantitative experimental design following the Cross-Industry Standard Process for Data Mining (CRISP-DM) framework. The data were obtained through observations and the collection of child health records at the Posyandu in Lubuk Sahung Village, Seluma Regency, Bengkulu Province, complemented by a public dataset from Kaggle and body images of children under five. The research process included data preprocessing, image feature extraction using CNN, feature fusion of health data and image features, model training using SVM and Random Forest, and model integration through Ensemble Voting. The results demonstrated that the SVM and Ensemble Voting models achieved the best predictive performance, with an accuracy of 91.67%, precision of 100.00%, recall of 84.62%, and an F1-score of 91.67%, while the Random Forest model achieved an accuracy of 83.33%. Based on the generalization analysis, the SVM model exhibited the highest stability, with a training–testing accuracy gap of only 4.16%. These findings indicate that integrating health data with image-derived features can effectively support the prediction of stunting levels and serve as a reliable decision-support tool for the early monitoring of children's nutritional status.

Analysis Of The Image Quality Of Public Works And Housing Infrastructure Documentation Using The Discrete Cosine Transform Method

Belly Prayoga, Devi Sartika, Ockhy Jey Fhiter — Mon, 26 Oct 2026 23:51:23 +0700

Visual documentation of Public Works and Public Housing (PUPR) infrastructure projects produces high-resolution images that function as progress evidence, technical verification, and audit archives; however, large file sizes create storage and transmission constraints. This study aims to analyze the quality of infrastructure documentation images using a web-based Discrete Cosine Transform (DCT) approach. The research employed a quantitative experimental design supported by the Waterfall development model. Infrastructure documentation images were processed through DCT transformation, quantization, reconstruction, and evaluation using Mean Squared Error (MSE), Peak Signal-to-Noise Ratio (PSNR), compression ratio, and processing time. The system was implemented using PHP and MySQL. Testing on five images produced an average MSE of 45.29 and an average PSNR of 34.86 dB, indicating that reconstructed images remained acceptable for visual documentation because most PSNR values exceeded 30 dB. Nevertheless, the average compression ratio of -68.98% indicated data expansion under the current implementation. These findings confirm that DCT is effective in preserving visual quality, but quantization parameters, output encoding strategy, and file-format handling require further optimization to achieve actual compression efficiency.

Performance Analysis of Discrete Cosine Transform and Discrete Wavelet Transform Methods on Digital Images Based on Peak Signal-to-Noise Ratio and Mean Squared Error Parameters

Muetya Oktaviani, Juju Jumadi, Ockhy Jey Fhiter. W — Mon, 26 Oct 2026 23:51:43 +0700

As the demand for image data storage and distribution continues to increase, compression techniques are required to reduce file size without significantly degrading image quality. One of the most widely used approaches in image compression is frequency-domain transformation, including the Discrete Cosine Transform (DCT) and the Discrete Wavelet Transform (DWT). This study aims to evaluate and compare the performance of DCT and DWT methods on digital images based on the Mean Squared Error (MSE) and Peak Signal-to-Noise Ratio (PSNR) parameters. In addition, the study also considers the compression ratio as a supporting parameter in the evaluation process. The results indicate that the DWT method produced an average MSE value of 63.9381, which was lower than that of DCT method at 70.0780. Meanwhile, DCT method achieved an average PSNR value of 31.21 dB, which was higher than that of DWT method at 30.52 dB. In terms of compression ratio, the DCT method also demonstrated better performance, with an average value of 4.2954, whereas DWT method produced an average value of 2.0275. Based on the testing results, it can be concluded that DWT method performs better in minimizing image distortion based on the MSE parameter, whereas the DCT method is superior in producing reconstructed image quality and compression efficiency based on PSNR values and compression ratio.

Comparative Design and Development of a Web-Based Wedding Organizer Booking Information System to Improve Efficiency and Service Quality of Wedding Organizers

Masdanar Alif Hidayatulloh, Sumarno Sumarno, Sukma Aji, Yulian Findawati — Mon, 26 Oct 2026 23:52:02 +0700

The development of information technology has encouraged the digitalization of various service sectors, including Wedding Organizer (WO) services. The ordering process on Amanda Wedding Organizer is still done manually, which has the potential to cause recording errors, schedule conflicts, delays in preparing reports, and limited access to information for customers. This research aims to design and build a web-based Wedding Organizer booking information system using the Django framework to improve operational efficiency and service quality. The research uses a qualitative descriptive approach with the Waterfall development model. Data was collected through observations, interviews with the owner and administrator of Wedding Organizer Amanda, and literature study, then analyzed to identify the functional and non-functional needs that underpinned the design of the system using the Unified Modeling Language (UML). The results of the research resulted in a web-based information system that provides package management features, online booking, payment uploads and verification, meeting schedule arrangements, order status monitoring, and the preparation of administrative reports through an integrated dashboard. Implementation of the Django framework supports more structured data management, minimizes the potential for administrative errors, and improves service access for customers and administrators alike. Thus, the system developed is able to support the digitization of Amanda's Wedding Organizer business processes so that operations become more effective, efficient, and documented in an integrated manner.

Design And Construction Of A Decision Support System For KIP Scholarship Selection Using The Multi-Attribute Utility Theory (MAUT) Method At SDN Sempur Purwakarta

Muhammad Fazri Sholihin, Mutiara Andayani Komara, Imay Kurniawan — Mon, 26 Oct 2026 23:52:27 +0700

The process of selecting recipients for the Kartu Indonesia Pintar (KIP) scholarship at SDN Sempur Purwakarta is still carried out manually, which requires a considerable amount of time and may lead to inaccuracies in decision-making. In addition, the uncomputerized assessment process causes difficulties in data processing and systematic student record management. Therefore, a system is needed to assist the selection process in a more objective and efficient manner. This study aims to design and develop a web-based Decision Support System (DSS) for selecting KIP scholarship recipients using the Multi attribute utility theory (MAUT) method. The system was developed using the Extreme programming (XP) method, which consists of planning, design, coding, and testing stages. The implementation of the system uses the PHP programming language with the Codeigniter 3 framework and MySQL database. The results of the study show that the MAUT method is able to produce a ranking process of scholarship candidates based on predefined criteria such as economic condition, number of dependents, and student attendance. The developed system provides recommendations more quickly, objectively, and systematically. Acceptance Testing results indicate that all system functions run properly, while usability evaluation using the System usability scale (SUS) achieved a high score, indicating that the system is easy to use and suitable for implementation in the school environment.

Application Of The Apriori Algorithm In The Point Of Sale (POS) System For Analyzing Consumer Purchase Patterns At Rahayu Motor Workshop

An Najm Reyhan, Ismi Kaniawulan, Mochzen Gito Resmi — Mon, 26 Oct 2026 23:53:01 +0700

Rahayu Motor Workshop faces sales optimization constraints due to unintegrated daily transaction data that merely accumulates as conventional archives. This lack of insight prevents the manager from predicting consumer purchasing patterns, leading to inventory overstock or stockouts. This study aims to develop a web-based Point of Sale (POS) system integrated with a data mining module using the Apriori Algorithm to identify item correlations. The system was built using the Rapid Application Development (RAD) method, utilizing CodeIgniter 4 and MySQL. The core engine employs Association Rule techniques to calculate support and confidence metrics from historical transaction histories. Functional evaluations via Black Box Testing confirmed that all modules operate validly. During the cutover phase, the system precisely extracted customer purchasing habits. This analytical output delivers actionable business intelligence for the workshop owner to design effective product bundling promotions, optimize shelf layouts, and streamline spare parts inventory control.

The Implementation Of Customer Satisfaction Assessment On Wira Computer Store Services Using The Service Quality (Servqual) Method

Tota Novriawan, Liza Yulianti, Veny Novita Sari — Mon, 26 Oct 2026 23:53:22 +0700

Customer satisfaction is one of the key factors that determines a company's success in maintaining customer loyalty. Wira Computer Store, as a provider of computer products and services, needs to assess customer satisfaction with the quality of its services in order to continuously improve service performance. The primary issue faced by the company is the absence of a system capable of objectively measuring customer satisfaction, resulting in service evaluations being conducted manually and inefficiently. Therefore, this study aimed to implement the Service Quality (SERVQUAL) method to measure customer satisfaction with the services provided by Wira Computer Store. The SERVQUAL method was applied by comparing customers' perceptions of the services received with their expectations across the five dimensions of service quality: Tangibles, Reliability, Responsiveness, Assurance, and Empathy. Data were collected through questionnaires distributed to customers and processed using a web-based information system developed with the PHP programming language, the Laravel framework, and the MySQL database management system. The results demonstrated that the system was able to calculate the service quality gap for each SERVQUAL dimension, enabling the identification of overall customer satisfaction levels as well as the specific service aspects requiring improvement. The implementation of this system allows Wira Computer Store to evaluate service quality more quickly, accurately, and systematically, thereby providing a reliable basis for continuous service quality improvement and enhancing customer satisfaction.

Decision Support System For Choosing The Best Layer Chicken Breed Using The Simple Additive Weighting (SAW) Method In Pasawahan Village,Purwakarta Regency

Naya Salsabilah, Mochzen Gito Resmi, Dede Irmayanti — Mon, 26 Oct 2026 23:53:54 +0700

Selecting a laying-hen breed is a critical decision because breeds differ in egg productivity, feed efficiency, disease resistance, environmental adaptability, age at first lay, egg quality, and mortality. In Pasawahan Village, Purwakarta Regency, selection has largely relied on experience and informal recommendations, making the assessment subjective. This study develops a decision support system that produces an objective recommendation using the Simple Additive Weighting (SAW) method. Data were collected through observation, interviews with farmers, and literature review. The system was developed using the Waterfall model, designed with UML, implemented with Flutter and Dart, and supported by Supabase as the database. Five alternatives were evaluated against seven weighted criteria, normalized, and aggregated into preference scores. Lohmann Brown achieved the highest score (0.871667), followed by Hy-Line Brown (0.851667), Isa Brown (0.846667), White Leghorn (0.740000), and KUB chicken (0.643333). Black-box testing of authentication, data management, SAW calculation, ranking, and user management functions was successful. The system makes breed selection more transparent, consistent, and structured; nevertheless, the criteria weights and sub-criteria scales should be validated across a broader farming population.

Design And Development Of A Progressive Web App (PWA)-Based Inventory Management System Using EOQ And ROP Algorithms

Mon, 26 Oct 2026 23:54:20 +0700

Advancements in information technology are driving the retail sector to move away from conventional systems in favor of improved operational efficiency. Pet Shop 4yos Veterinary Care faces challenges regarding manual inventory tracking, which leads to risks such as stock discrepancies and procurement delays. This study aims to integrate the Economic Order Quantity (EOQ) and Reorder Point (ROP) algorithms into a Progressive Web App (PWA)-based information system. The PWA method was selected for its ability to provide responsive, multi-platform access, thereby facilitating easier inventory monitoring. Application of the mathematical formulas indicates that the ideal economic order quantity for the shop is 154 units per procurement cycle. The system also successfully generates accurate, automated reorder notifications precisely when warehouse stock levels drop to 41 units. Evaluation results demonstrate that the application functions effectively and maintains accurate, real-time stock data updates. The primary contribution of this research is the provision of an effective and practical digital warehouse management system model designed to minimize the risk of inventory mismanagement in the business.

Implementation of The Organization Rangement Synthese Dedonnes Relationnelles (ORESTE) Method in Determining The Quality of Oyster Mushrooms Suitable For Marketing

Dinta Putri Nabila, Dewi Suranti, Ila Yati Beti — Mon, 26 Oct 2026 23:54:46 +0700

Oyster mushrooms are a horticultural commodity with high economic value and are highly sought after by the public. However, determining the quality of oyster mushrooms suitable for marketing is often done subjectively, potentially leading to inconsistencies in decision-making. Therefore, this study aims to implement the Organization Rangement Synthese Dedonnes Relationnelles (ORESTE) method to objectively and systematically determine the quality of oyster mushrooms suitable for marketing. The ORESTE method is a Multi-Criteria Decision Making (MCDM) method capable of ranking alternatives based on preference relationships between criteria without requiring data normalization. The criteria used in this study included color, size, texture, freshness, and physical condition of oyster mushrooms. The results showed that the ORESTE method can provide an accurate ranking of oyster mushroom quality and assist business owners in determining marketable oyster mushrooms. Therefore, the implementation of the ORESTE method is expected to be an effective decision support solution for improving the quality and competitiveness of oyster mushroom products in the market.

Rapid Application Development (RAD) in Creating Flutter Applications for Sales Monitoring

Bintang Andika Adirizki, Ujang Juhardi — Mon, 26 Oct 2026 23:55:07 +0700

Digitalization in small retail businesses is required to reduce recording errors, accelerate report preparation, and provide real-time sales information. This study aimed to design and develop a mobile application for sales monitoring and inventory management using Flutter and Firebase Firestore, and to evaluate its functional feasibility and user acceptance. The system was developed using the Rapid Application Development (RAD) approach through Planning, User Design, Construction, and Cutover phases. Application requirements were identified through observation and interviews at a grocery store in Bengkulu City. The resulting system provides user authentication, a real-time monitoring dashboard, inventory management, incoming-goods and sales transaction recording, and PDF sales-report generation. Black Box Testing was conducted on 12 scenarios, all of which were successfully completed. User Acceptance Testing involved 23 respondents and eight Likert-scale statements; the application obtained a score of 832 out of 920, equivalent to 90.43%, and was classified as very good. These findings indicate that the application performs its main functions reliably, replaces inefficient manual recording, and is well accepted as a tool to support grocery-store operations and business decision-making.

Decision Support System For Determining Toddler Nutrition Status Using Web-Based Simple Additive Weighting Method (Case Study: UPTD Tegal Timur Health Center)

Adi Naufal Fatchurohman, Slamet Wiyono, Syefudin Syefudin — Mon, 26 Oct 2026 23:55:29 +0700

The determination of nutritional status and growth of toddlers at UPTD Puskesmas Tegal Timur still relies on manual anthropometric calculation based on the Ministry of Health's standards, which requires high precision, takes relatively long, and does not yet support structured prioritization of toddler treatment. This study develops a web-based decision support system using the Simple Additive Weighting (SAW) method to assist the process of determining toddlers' nutritional status and growth. The system is built using the Laravel framework and MySQL database, integrating Z-Score calculation based on Indonesia's Child Anthropometric Standard (Ministry of Health Regulation No. 2 of 2020) into the SAW weighting process. Z-Score values from anthropometric indicators form the basis of the decision matrix, normalization, preference value calculation, and toddler priority ranking. The system implementation provides features for managing toddler data, assessment data, criteria and weights, SAW calculation, growth chart visualization, and reporting. Testing was conducted using the Black Box method and validated by comparing system results with manual calculation. All system functions worked as required, and calculation results at every stage were consistent with manual calculation, so the system provides a more objective basis for determining toddler treatment priority.

Analysis And Planning Of Enterprise Architecture In Developing An Application Ecosystem Using The Zachman Framework At PT Nusantara Sukses Teknologi

Moh Julhan Arif, Iffatul Mardhiyah, Rudi Hartono — Mon, 26 Oct 2026 23:55:50 +0700

The development of integrated systems plays an important role in enhancing business opportunities, optimizing operational processes, and maximizing information delivery to users. System integration is necessary to connect applications operating in different environments to support effective and structured service management. This study aims to design a blueprint for developing the Nusatek application ecosystem by implementing the Zachman Framework as an information system architecture approach. The Nusatek ecosystem includes applications for daily service needs and certification services operating in different system environments. The Zachman Framework is applied to map requirements, components, processes, and relationships between systems as a foundation for application integration planning. The design is visualized through a mini prototype illustrating the conceptual connectivity between applications within an integrated ecosystem. This proposed architecture is expected to guide systematic development, improve service management efficiency, strengthen information delivery, and support sustainable business opportunities through information technology solutions aligned with organizational needs and future development priorities.

Air Pollucion Level Estimacion In Public Areas Using Image Segmentation And Validation Against Air Quality Sensor Data

Muhamad Reypi, Rozali Toyib, Yulia Darmi, Yopi Apridiansyah, Yusa Virginiawan — Mon, 26 Oct 2026 23:56:12 +0700

This study develops an air pollution level estimation system based on digital image processing, integrated with validation using a CO₂ sensor, and implemented as a Python-based desktop application. The proposed method employs image segmentation in the HSV color space using a multi-zone thresholding approach to detect smoke, haze, and other visual pollutants. Feature extraction is then performed by calculating the pollution area ratio, texture entropy, mean intensity, standard deviation, and average saturation. These features are combined into a composite score that serves as the basis for classifying air pollution levels into three categories: Low, Moderate, and High. In parallel, the system analyzes images of the CO2 sensor display by detecting the color ratio of green, yellow, and red indicators to estimate atmospheric CO₂ concentration levels. Cross-validation between the two methods was conducted using 100 paired images to evaluate the consistency of classification results. Evaluation results based on a confusion matrix indicate an overall system accuracy of 70.00%, with a Macro Precision of 37.90%, Macro Recall of 45.93%, and Macro F1-Score of 36.70%. The High category demonstrated the best performance, achieving an F1-Score of 81.53%, while the Moderate category obtained an F1-Score of 28.57%. The Low category could not be effectively detected due to data distribution imbalance. Cross-validation results showed a 37% agreement rate between image segmentation and CO2 sensor measurements, indicating that further improvements are required. This result also reflects the fundamental difference between visual pollution indicators and direct gas concentration measurements. The developed system is equipped with an interactive graphical user interface, confusion matrix visualization, and a multi-sheet Excel report export feature that includes detailed results, performance evaluation, sensor analysis, as well as automatically generated conclusions and recommendations.

Developing An Android-Based Point Of Sale (Pos) Application For Warung Hadi Using Firebase

Andrian Kuor Nando, Harry Witriyono, Ardi Wijaya, Agung Kharisma Hidayah — Mon, 26 Oct 2026 23:56:31 +0700

The rapid development of information technology encourages small businesses, including traditional shops, to shift from manual systems to digital systems in order to improve efficiency and accuracy in business management. Warung Hadi, as the object of this study, still uses manual record-keeping, which is prone to errors such as miscalculations, recording mistakes, and stock management issues. Therefore, this study aims to design and develop an Android-based Point of Sale (POS) application integrated with Firebase as the backend service.

The research methods include observation, interviews, and literature study for data collection, as well as system development stages consisting of analysis, design, implementation, and testing. The application was

developed using the Kotlin programming language with Android Studio as the development environment, and Firebase for real-time database management.

The results show that the developed POS application includes main features such as product management, sales transactions, and sales reports, which help improve the operational efficiency of Warung Hadi. Based on Black Box Testing, all application features function properly with a success rate of 100%.

In conclusion, this application provides an effective digital solution to overcome manual recording problems and supports more structured, faster, and more accurate business management.

Whatsapp Notifications For Soil Moisture Information In An Automatic Plant Watering System Using A Soil Moisture Sensor

Yuza Reswan, Genta Dedi Kumara, Dandi Sunardi, Harry Witriyono — Tue, 27 Oct 2026 11:20:48 +0700

Manual plant watering has several limitations because soil moisture conditions cannot be continuously monitored, which may result in delayed or excessive watering. These conditions can inhibit plant growth because each plant requires an appropriate amount of water based on its soil moisture level. Therefore, an automatic plant watering system capable of operating automatically and providing real-time information on soil moisture conditions is required. This study aims to design an automatic plant watering system using a soil moisture sensor and ESP32 with WhatsApp as a notification medium for soil moisture conditions. The research employed the prototype method, which consisted of literature review, system design, system testing, data analysis, and conclusion drawing. The developed system utilizes a soil moisture sensor to detect soil moisture levels and an ESP32 microcontroller to process sensor data and automatically control the water pump. In addition, WhatsApp is used as a notification service to deliver information to users. The system operates by periodically reading soil moisture values using the soil moisture sensor. The collected data are then processed by the ESP32 to determine the soil condition based on a predefined threshold value. When the soil is detected as dry, the water pump is automatically activated to irrigate the plants. Conversely, when the soil reaches the required moisture level, the water pump is automatically turned off. Furthermore, the system sends notifications via WhatsApp, allowing users to monitor soil moisture conditions and the watering status without directly inspecting the plants. The results showed that the system was able to distinguish between dry and wet soil conditions using a threshold value of 2500. The sensor reading for dry soil was 4095, while the reading for wet soil was 1383. The system also successfully controlled the water pump automatically and sent WhatsApp notifications with an average delay of approximately 5 seconds. Based on the testing results, the system was able to monitor soil moisture and provide users with information through WhatsApp.

A Sentiment Analysis On Youtube Comments On Sumatera Flash Flood Disaster Videos Using The Long Short-Term Memory (LSTM) Method

Adithyo Ramadhan, Harry Witriyono, Muntahanah Muntahanah, Muhammad Husni Rifqo — Tue, 27 Oct 2026 11:21:08 +0700

The flash flood disaster in Sumatra generated various public responses, many of which were expressed through comments on the YouTube platform. These comments contain opinions, criticism, support, and emotional expressions that can be analyzed to identify public sentiment trends regarding the disaster. This study aims to analyze the sentiment of YouTube comments related to flash flood disaster videos in Sumatra using the Long Short-Term Memory (LSTM) method. The research data were collected through the YouTube Data API v3, resulting in 6,097 Indonesian-language comments. The research stages included data collection, preprocessing consisting of cleaning, case folding, word normalization, tokenization, stopword removal, and stemming, sentiment labeling using a lexicon-based approach, data splitting with a ratio of 80% training data and 20% testing data, and sentiment classification using an LSTM model. The labeling results showed that negative sentiment dominated with 2,234 comments (37.69%), followed by positive sentiment with 2,011 comments (33.93%), and neutral sentiment with 1,682 comments (28.38%). Based on the evaluation results, the LSTM model achieved an accuracy of 77.64%, precision of 78.26%, recall of 77.55%, and an F1-score of 77.45%. The findings indicate that the Long Short-Term Memory (LSTM) method is capable of classifying YouTube comment sentiments effectively and can be utilized to represent public opinion regarding flash flood disasters in an automated and objective manner.

Implementation Of Morphological Skeletonization For Leaf Vein Extraction From Leaf Images

Muhamad Fadhili, Gunawan Gunawan, Ardi Wijaya, Yovi Apridiansyah — Tue, 27 Oct 2026 11:21:26 +0700

This study implements the Morphological Skeletonization method to automatically extract leaf vein structures from mango leaf (Mangifera indica) images. The developed system utilizes an image processing pipeline that includes HSV-based leaf segmentation, preprocessing with Bilateral Filter, vein contrast enhancement using dual-scale Black-Hat Transform and Contrast Limited Adaptive Histogram Equalization (CLAHE), binarization using Otsu and Adaptive Thresholding methods, morphological opening and closing operations, and skeletonization using the iterative thinning algorithm from the scikit-image library. Performance evaluation was conducted quantitatively using Mean Squared Error (MSE) metrics at two stages, namely Binarization MSE and Skeleton MSE. Testing was performed on a dataset consisting of 150 mango leaf images in JPG format collected using a smartphone camera under standardized acquisition conditions. The results show that the system achieves an average Skeleton accuracy of 94.47% with a standard deviation of 1.06%, an average Binary accuracy of 61.28%, and an average Global accuracy of 77.87%. The average Binarization MSE of 25179.25 and Skeleton MSE of 3599.02 confirm that the system is capable of extracting leaf vein skeletons consistently and accurately across the entire dataset. The extracted skeleton structural features include length, number of branch points, connectivity, and density, which quantitatively represent the morphological characteristics of the leaf venation network. This study demonstrates that the mathematical morphology approach offers an efficient, transparent, and interpretable solution for leaf vein extraction tasks compared to deep learning-based approaches that require large computational resources.

Machine Learning Analysis Using the Hybrid K-Means and C4.5 Model in Determining Student Readiness for Educational Field Practice

Nilam Agsera, Musli Yanto, Syafri Arlis — Tue, 27 Oct 2026 11:21:45 +0700

The evaluation of student readiness for Educational Field Practice (PLK) has conventionally relied on the subjective assessments of supervising lecturers and mentor teachers, posing a risk of inaccurate outcomes. This traditional framework continues to depend on manual, non-standardized analytical approaches. To address these limitations, this study aims to develop a hybrid K-Means and C4.5 machine learning model to categorize the readiness levels of PLK students at Universitas Negeri Padang. K-Means clustering was initially applied to group evaluation data into three distinct readiness categories, with the resulting cluster labels serving as target variables. Subsequently, the C4.5 algorithm was utilized to extract decision rules and generate a predictive decision tree. The dataset comprises 3,578 student grade records obtained from the Center for Field Experience (Pusat PPL) at Universitas Negeri Padang for the July–December 2025 period, encompassing ten assessment indicator variables ranging from partner school observations to final practice reports. K-Means clustering classified 2,619 students (73.20%) into the "Ready" category, 936 students (26.16%) as "Sufficiently Ready," and 23 students (0.64%) as "Needs Mentoring". The C4.5 model selected the learning media indicator (X4) as the root node, achieving an overall accuracy of 92.32%, a weighted precision of 92.33%, a recall of 92.32%, and an F1-score of 92.33%. These findings demonstrate that the hybrid K-Means–C4.5 approach provides an objective, transparent, and data-driven instrument to support PLK administrators in final evaluations and in determining student readiness status upon completing field practice.

Analysis Of Closing Price Forecasting For Bank Negara Indonesia Stocks Using A Hybrid K-Means Backpropagation Neural Network Model

Alfajri Salim, Sarjon Defit, Gunadi Widi Nurcahyo — Tue, 27 Oct 2026 11:22:03 +0700

Stock price movements are dynamic in nature, and their active patterns make them difficult to predict accurately. These conditions have driven the development of forecasting models capable of recognizing data characteristics before generating predictions with a low error rate. This study aims to analyze how the hybrid K-Means and Backpropagation Neural Network model works in forecasting stock closing prices. The dataset used consists of historical closing prices of Bank Negara Indonesia (BNI) stock over a 5-year period, comprising a total of 1,447 data points. Based on the results of this study, the model achieved an RMSE of 109.26, an MAE of 83.30, and a MAPE of 2.1458%, indicating that the model is capable of predicting stock prices with a low error rate based on these evaluation metrics. The addition of klasters to the neural network training process helps the model recognize data characteristics, resulting in more stable predictions. This study provides a hybrid-model-based approach to stock closing price forecasting that can serve as a reference for developing time-series prediction models in the field of finance.

Analysis of Student Performance Determination Using K-Means Clustering and Perceptron at Private Madrasah Tsanawiyah Koto Anau

Zakiya Hikmi, Musli Yanto, Syafri Arlis — Tue, 27 Oct 2026 11:22:20 +0700

The way we normally evaluate how well students are doing at MTsS Koto Anau is not good enough. This is because it does not show us the hidden patterns of how students learn. This study wants to use a way to evaluate how well students are doing. It uses K-Means Clustering and Perceptron together to get an idea of what helps students achieve. We used a two step method. First K-Means Clustering puts students into three groups based on how similar they're in terms of school work. Then we use these groups to teach the Perceptron model. The information we used came from the office at MTsS Koto Anau. It had details about 113 students. Included things like their average grades how often they came to school and their final exam scores. When we tested the model it was very good at putting students into the group. It was 99 percent of the time. This shows that the model is very good, at figuring out which group a student belongs to. The main thing we found out is that using K-Means Clustering and Perceptron is a good way to understand how students learn and predict how well new students will do. This study helps because it gives madrasahs an useful way to analyze student data. This means they can make learning plans that're just right for each student. The MTsS Koto Anau can use this to make sure students get the help they need to do their best.

Analysis of Student Interest Recommendations Using K-Means and Perceptron Algorithm at MTSN 3 Solok

Hafis Hidayattullah, Rini Sovia, Larissa Navia Rani — Tue, 27 Oct 2026 13:30:10 +0700

Choosing an academic field before entering Vocational High School (SMK) should be supported by information reflecting students' abilities. This study aims to provide an initial field recommendation using K-Means Clustering and Single Layer Perceptron (SLP). K-Means forms three levels of ability, namely High, Medium, and Low, based on nine subject grades. SLP identifies tendencies toward Science, Social, or Creative fields using report card grades and the difference between the two highest field averages. The dataset contains 1,000 records from the 2021/2022 to 2024/2025 academic years. K-Means produced a Silhouette Score of 0.6734 and converged at the seventh iteration. On 150 testing records, SLP achieved 97.33% accuracy, 97.36% precision, 97.33% recall, and a 97.34% F1-score, with 146 correct predictions. Both outputs were combined into an initial recommendation containing ability level and field tendency for students and schools, rather than a final decision. Final selection should consider interests, talents, conditions, and teacher guidance

Comparison of K-Nearest Neighbor and Naive Bayes Algorithms for Vocational High School 4 Padang Student’s Majoring Specialization

Ryan Firmansyah, Agung Ramadhanu, Billy Hendrik — Tue, 27 Oct 2026 13:30:37 +0700

Determining a student's major of specialization at State Vocational High School (SMKN) Negeri 4 Padang is still done subjectively, so there is a risk of causing inaccurate placement which has an impact on student learning achievement. A Data Mining approach is needed to analyze academic grade patterns in a structured and objective manner to support majoring decisions. This research aims to implement and compare the performance of the K-Nearest Neighbor (KNN) algorithm and the Naive Bayes Classifier in predicting students' majors. The research method applies a standard preprocessing approach, which includes data cleaning, normalization, and training-test data sharing. The classification model was then evaluated using the Stratified 5-Fold Cross Validation and Confusion Matrix methods to assess the accuracy and reliability of predictions. The dataset consists of 68 records of SMKN 4 Padang students with five core subject variables and three target major specialization classes. Experimental results show that the Naive Bayes Classifier produces superior performance with an average total accuracy of 64.71%. Meanwhile, the K-Nearest Neighbor algorithm (K=12) obtained an accuracy of 47.06%. The contribution of this research can provide an automated decision support system model based on objective academic data for schools in determining major interest.

Sentiment Analysis Of Public Reviews Of Public Services At The Dpmptsp Of Bengkulu Province Using The Naive Bayes Method

Tue, 27 Oct 2026 13:32:01 +0700

This study aims to analyze public sentiment toward public services at DPMPTSP of Bengkulu Province using the Naive Bayes method. The research data were obtained from public reviews collected during the 2025–2026 period and processed using Natural Language Processing (NLP) techniques, including case folding, cleaning, tokenizing, stopword removal, and stemming. Furthermore, the data were represented using the TF-IDF method and classified into positive, negative, and neutral sentiments using the Naive Bayes algorithm. The results indicate that the Naive Bayes method is capable of classifying public review sentiments effectively and providing an overview of public satisfaction and perceptions regarding the services provided. The findings of this study are expected to serve as evaluation material for DPMPTSP of Bengkulu Province in its efforts to improve the quality of public services.

Analysis Of The Performance Of The Wifi Network At The Faculty Of Engineering, Muhammadiyah University Of Bengkulu, Using Qos (Quality Of Service) Parameters

Tue, 27 Oct 2026 13:32:21 +0700

The quality of the Wi-Fi network plays an important role in supporting academic activities at the Faculty of Engineering, Muhammadiyah University of Bengkulu. This study aims to analyze Wi-Fi network performance using the Quality of Service (QoS) method based on parameters like throughput, packet loss, delay, and jitter. Measurements were taken on six access points located on the 1st, 2nd, and 3rd floors during busy and quiet hours using the Wireshark application. The measurement results were analyzed and compared with TIPHON standards to determine network quality. The research results showed that the average throughput value of 770.19 Kbps falls into the Fair category, packet loss at 0.01% is in the Very Good category, delay of 32.49 ms is in the Very Good category, and jitter of 32.65 ms is in the Good category. Based on this analysis, an average index of 3.25 was obtained, so the overall Wi-Fi quality at the Faculty of Engineering, Muhammadiyah University of Bengkulu is in the Good category according to TIPHON standards and can support academic activities.

The Effect of Information Quality and Service Quality on User Satisfaction with the Public Service Information System at the National Unity and Political Affairs Agency of Palembang City

Poni Cornelia, Tri Oktarina — Tue, 27 Oct 2026 13:32:38 +0700

Digitalizing public services does not automatically ensure that a service will be adopted or evaluated satisfactorily. This study examines the effects of information quality and service quality on user satisfaction with the Public Service Information System of the National Unity and Politics Agency of Palembang City by adapting selected constructs from the DeLone and McLean model. A cross-sectional quantitative survey involved 141 users purposively selected from a population of 218 users. The instrument comprised 14 Likert-scale items measuring information quality, service quality, and user satisfaction. The data were analyzed using PLS-SEM through measurement and structural model assessment. All indicator loadings ranged from 0.736 to 0.858; AVE values ranged from 0.585 to 0.663; and composite reliability ranged from 0.849 to 0.908. Information quality had a positive and significant effect on user satisfaction (beta=0.442; t=7.100; p<0.001), as did service quality (beta=0.460; t=8.777; p<0.001). Together, the two constructs explained 44.6% of the variance in user satisfaction. The results identify regularly updated service information and responsive user support as improvement priorities. The findings are limited to people who had used the system and do not represent every success dimension in the DeLone and McLean model.

Implementation of the Random Forest Regression Algorithm for Estimating the Number of Personnel at the Tangerang City Fire Department Call Center

Rahel Azzahra, Kahfi Heryandi Suradiradja — Tue, 27 Oct 2026 13:32:56 +0700

This study aims to apply the Random Forest Regression algorithm to estimate the number of personnel required for the Tangerang City Fire Department Call Center, based on historical fire incident data from 2023 to 2025. The research addresses the need for rapid, accurate, and data-driven personnel estimation to support decision-making by personnel. The study employs the Cross-Industry Standard Process for Data Mining (CRISP-DM) methodology, comprising Business Understanding, Data Understanding, Data Preparation, Modeling, Evaluation, and Understanding. The variables utilized include year, time of incident, district, object type, fire severity level, weather, fleet size, and response time, with the number of personnel serving as the target variable. The study compares Random Forest Regression with Linear Regression using MAE, MSE, RMSE, and R² metrics. Test results indicate that Random Forest Regression achieved an MAE of 1.976, MSE of 6.085, RMSE of 2.467, and R² of 0.953, demonstrating superior performance compared to Linear Regression. The model was subsequently implemented in a Flask-based web application to serve as an operational tool for relevant personnel

The Application Of The Transfer Learning Method In The Classification Of Eggplant Diseases Using The Densenet-201 Architecture

Hajar Rojandra, Sastya Hendri Wibowo, Erwin Dwika Putra, Harry Witriyono — Tue, 27 Oct 2026 13:33:30 +0700

Disease affecting eggplant fruit (Solanum melongena L.) is one of the main obstacles that reduce crop productivity, while conventional detection methods are considered inefficient as they require expert knowledge and a considerable amount of time. This study aims to apply the transfer learning method with the DenseNet201 architecture to classify eggplant fruit diseases into three categories, namely healthy fruit, rotten fruit, and fruit borer, based on digital images. The research dataset consisted of 2,417 images, divided into training, validation, and testing data with a 70:15:15 ratio, which were then processed through resizing, normalization, denoising, and data augmentation on the training set so that each class contained 1,000 images. The model was built using pre-trained ImageNet weights and trained for 20 epochs with the Adam optimizer, achieving a training accuracy of 90.0% and a validation accuracy of 87.5% with no significant indication of overfitting. During testing, the model achieved a test accuracy of 81.69% with a weighted F1-score of 81.68%. The healthy fruit class showed the best performance with an F1-score of 90.77%, while the rotten fruit and fruit borer classes achieved F1-scores of 78.48% and 74.11%, respectively, due to the visual similarity between the two classes. System testing on sample images from each class resulted in entirely accurate classifications, with confidence values ranging from 78.35% to 99.69%. These results indicate that the DenseNet201 model based on transfer learning performs well and has the potential to be used as an early detection support system for eggplant fruit disease.

Moving Object Detection Using The Yolo Method And Video-Based Depth Estimation

Yuza Reswan, Erwin Dwika Putra, Yovi Apridiansyah, Dwi Bagas Prakoso — Tue, 27 Oct 2026 13:33:50 +0700

The development of artificial intelligence (Artificial Intelligence), particularly in the field of computer vision, has driven advances in image and video processing to automatically understand environmental conditions. One of its applications is object detection; however, detection methods generally only provide information about the type and location of objects within bounding boxes without information about their distance from the camera. This limitation means that the information generated is insufficient to support systems requiring a more comprehensive understanding of the environment. Therefore, an approach that combines object detection and distance estimation using an RGB camera without additional hardware is required. This study aims to develop a video-based object detection and distance estimation system using a combination of YOLOv10 and ZoeDepth. YOLOv10 is used to detect objects and determine their positions in each video frame, while ZoeDepth is used to perform depth estimation to obtain the estimated distance of objects from the camera. The research methodology includes data collection, preprocessing, model implementation, and system evaluation. The data consist of 250 CCTV video samples, comprising 162 secondary data and 88 primary data. All data were used as testing data because YOLOv10 is a pre-trained model previously trained using the COCO dataset. Object detection performance was evaluated using the mAP metric, while distance estimation was evaluated using Root Mean Square Error (RMSE), with the iBims-1 dataset serving as the ground-truth depth reference. The test results show that the system is capable of detecting several objects, namely cars, buses, trucks, and persons, while also providing distance estimates for the detected objects. The object detection evaluation achieved an mAP value of 81.03%, while the distance estimation evaluation using ZoeDepth produced an RMSE value of 4.2248 meters. Based on these results, the combination of YOLOv10 and ZoeDepth successfully produces a system capable of performing object detection while simultaneously providing video-based distance estimation using an RGB camera. This system can serve as an approach for developing computer vision applications that require both object and distance information. This approach can improve the completeness of visual information to support automated environmental monitoring.

Comparative Analysis of Moving Average and Single Exponential Smoothing Methods for Optimizing Inventory Forecasting

Rizki Ramadhon Pratama, Lena Elfianty, Fahrul Ikram Nizar — Tue, 27 Oct 2026 13:34:17 +0700

Inventory management in small retail businesses is often affected by demand uncertainty because sales recording and stock monitoring are still performed using simple procedures. This condition occurs at Toko Ipung and can lead to both overstock and stockout situations. This study aims to implement and compare the Moving Average and Single Exponential Smoothing methods to produce more measurable inventory forecasts. A quantitative comparative approach was combined with web-based system development using the Waterfall model. The main data consist of historical sales from January to December 2025; the manual calculation simulation presents 15 product types, while the overall evaluation in the study is reported for 60 product types. The system was developed using PHP and MySQL, with Moving Average and Single Exponential Smoothing using a smoothing constant of 0.2 as the forecasting algorithms. Forecast accuracy was evaluated using Mean Absolute Percentage Error, while application functions were verified through Black Box Testing. The results show that all functional test scenarios were successfully executed and that the overall average Mean Absolute Percentage Error was 15.14%, which falls into the good category. For the Beras Rojo Lele example, Moving Average produced 12.55% and Single Exponential Smoothing produced 12.50%, making the latter slightly more accurate. The system therefore provides an objective basis for procurement planning and for reducing stockout and overstock risks.

Application Of The Dbscan Algorithm For Clustering Criminal Evidence Data At The Bengkulu District Attorney's Office

Agung Biantoro Mahendra, Dewi Suranti, Prahasti Prahasti — Tue, 27 Oct 2026 13:34:36 +0700

Evidence management at the Bengkulu District Attorney's Office was previously handled manually, which complicated storage and retrieval as caseloads increased. This study implements the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm to group 150 evidence records based on evidence type, quantity, date received, and storage location. A quantitative-experimental design was used, covering data pre-processing, encoding, normalization, and clustering with Python on Google Colab, using epsilon = 0.8 and minimum points = 3. DBSCAN produced four clusters and identified eight records as noise, achieving a Silhouette Score of 0.4757, indicating moderate clustering quality. The clusters were respectively dominated by electronics, vehicles, sharp weapons, and documents, each linked to a separate storage warehouse. These findings show that DBSCAN can support more structured and efficient evidence storage and retrieval governance within the prosecutor's office.

Implementation Of Eye Tracking For Cheating Detection In Computer Vision-Based Exam Simulations

Yovi Apridiansyah, Gerry Mandala Putra, Gunawan Gunawan, Ardi Wijaya — Tue, 27 Oct 2026 15:03:56 +0700

This study implements an eye tracking system based on computer vision to detect cheating behavior during computer-based examination simulations. The developed system utilizes the MediaPipe Face Mesh library to detect more than 468 facial landmark points without requiring specialized hardware, then extracts visual behavior features including gaze direction, gaze-outside-screen duration, saccade count, and PERCLOS (Percentage of Eye Closure) values from examinee video recordings using OpenCV and Python through a Tkinter-based GUI. Cheating detection is performed using a rule-based threshold approach, without requiring machine learning model training. Testing was conducted on a dataset consisting of 30 examination simulation videos of approximately 10–20 seconds each, comprising 15 videos labeled as cheating and 15 videos labeled as normal ground truth. System performance was evaluated using a confusion matrix with the cheating class as the positive class. The results show that the system achieves an accuracy of 93.33%, precision of 100%, recall (sensitivity) of 86.67%, F1-score of 92.86%, and specificity of 100%, with True Positive of 13, True Negative of 15, False Positive of 0, and False Negative of 2. These results indicate that the PERCLOS value is the most dominant indicator in the cheating detection process in this study. This study demonstrates that a threshold-based eye tracking approach offers a lightweight, transparent, and affordable cheating detection solution compared to approaches requiring specialized hardware or deep learning models that require large computational resources.

Design Of A Talent And Interest Assessment (Abm) Test Application At Smkn 3 Bengkulu City Using The Fisher-Yates Shuffle Algorithm

Yuza Reswan, Muhammad Husni Rifqo, Khairullah Khairullah, Ade Rangga Saputera — Tue, 27 Oct 2026 15:04:13 +0700

In the world of education, especially at the Vocational High School (SMK) level, students are expected to be able to choose a major that suits their talents and interests so that the learning process can take place optimally and produce graduates who are competent in their fields. The development of web-based or computer-based applications for the implementation of assessment tests is an effective solution. Assessment test applications can provide convenience in presenting questions, randomizing the order of questions so that the results are more objective. However, in reality, there are still many students at vocational schools in Bengkulu City who do not know the direction and goals they want to achieve, especially in continuing their further studies. How to design and build an aptitude and interest assessment (ABM) test application by applying the Fisher-Yates Shuffle algorithm at SMK N 3 Bengkulu City.

Development Of A Bookkeeping And Financial Reporting System For A Waste Bank

Vepi Ega Lestari, Marhalim Marhalim, Ujang Juhardi, Pahrizal Pahrizal — Tue, 27 Oct 2026 15:04:30 +0700

This study aims to develop a Bookkeeping and Financial Reporting System for the Waste Bank at RT 02 RW 01, Kandang Limun Village. The main problems faced are the manual process of recording transactions and preparing financial reports, which may result in recording errors, delays in data recapitulation, and difficulties in managing and presenting data in a structured manner. This study uses a qualitative approach with data collection techniques consisting of observation, interviews, and literature review. The system development uses the Waterfall method, which consists of requirements analysis, system design, coding, testing, implementation, and maintenance. The developed system includes customer data management, waste type management, waste deposit transactions, withdrawal transactions, bookkeeping, and financial reporting. System testing is conducted using the Black Box Testing method to ensure that each system function operates according to user requirements. The results of this study are a bookkeeping and financial reporting system that can assist the waste bank administrators in recording transactions, managing data, and preparing financial reports more effectively, accurately, transparently, and systematically. The developed system is expected to improve the organization of administrative and financial management processes at the Waste Bank and reduce dependence on manual record-keeping

The Application of XGBoost to Classify Students’ Interests in Computer Science at SMP Negeri 30 in South Bengkulu

Ocy Lara Putri, Devi Sartika, Dimas Aulia Trianggana — Tue, 27 Oct 2026 15:04:48 +0700

Student interest in Informatics is still commonly identified through teacher observation, creating a need for a more objective data-driven approach. This study develops an XGBoost classification model and explains the factors underlying its decisions using SHAP. A quantitative data-analytics design followed the CRISP-DM stages using 27 student records and seven features: assignment scores, daily tests, practical scores, midterm scores, final scores, attendance, and attitude. The dataset was split into 80% training and 20% testing data. On six held-out cases, the model correctly classified two students as Interested and four as Not Interested, producing 100% Accuracy, Precision, Recall, and F1-Score, with an AUC of 1.00. SHAP analysis identified daily-test scores as the strongest contributor (1.1179), followed by assignment, final-exam, practical, and midterm scores. These findings indicate that XGBoost combined with SHAP can support measurable and transparent identification of student interest, although the limited dataset requires caution when generalizing the performance.

Implementation Of The Fcfs Algorithm In The Ari Barbershop Booking System Based On Time Slot Scheduling

M Shidiq Alfajar, Marhalim Marhalim, Yovi Apridiansyah, Ujang Juhardi — Tue, 27 Oct 2026 15:05:05 +0700

This research aims to design and build a web-based barbershop booking system at Ari Barbershop by implementing the First Come First Served (FCFS) algorithm combined with time slot scheduling. The main problem encountered is the manual queuing process, which often results in customer congestion, uncertain waiting times, and inefficient service. The solution offered is an online booking system that allows customers to make reservations in advance by selecting a service time (time slot). The FCFS algorithm is used to determine the order of service based on the booking time, thus ensuring fairness in the queue. The concept of time slot scheduling is used to divide service time to avoid schedule conflicts and minimize customer waiting time. The system development method used is the waterfall method, which includes the stages of requirements analysis, system design, implementation, and testing. This system was built using web technologies such as HTML, CSS, PHP, JavaScript, and MySQL. The expected outcome of this research is the creation of a system that can improve queue management efficiency, simplify the booking process for customers, and enhance service quality at Ari Barbershop.

Design and Development of a Web-Based Information System for Managing Infrastructure and Facility Complaints at the Audit Board of the Republic of Indonesia, North Sumatra Provincial Representative Office

Syafrizal Amanda Tambunan, Muhammad Zen, Hafni Hafni — Tue, 27 Oct 2026 15:05:23 +0700

Facilities and infrastructure complaint management at the North Sumatra Provincial Representative Office of the Audit Board of the Republic of Indonesia is still handled through verbal communication or WhatsApp, causing limited documentation, status monitoring, and report recapitulation. This condition may lead to delays in follow-up actions, unclear responsibilities, and limited evaluation data for internal services. This study aims to design and develop a web-based facilities and infrastructure complaint information system that supports complaint submission, verification, disposition, follow-up, status monitoring, user management, and report recapitulation. This research applies an applied research approach using the System Development Life Cycle Waterfall model, consisting of requirements analysis, design, implementation, testing, and maintenance. The system is designed based on four roles: user, operator, technician, and administrator. The technologies used include PHP with the Laravel Framework, Bootstrap 5, JavaScript, jQuery, DataTables, and MySQL/MariaDB. The results include a web-based system design and prototype, UML modeling, interface design, peran-based feature mapping, and functional testing scenarios using Black Box Testing. The system is expected to improve report traceability, clarify disposition workflows, accelerate follow-up actions, and support more transparent and accountable internal complaint governance.

Customer Sentiment Analysis Toward MSME Products On Tiktok Social Media Using Naive Bayes Algorithm

Muhammad Husni Rifqo, Elezah Syafitri, Khairullah Khairullah, Ujang Juhardi — Wed, 28 Oct 2026 02:10:32 +0700

This The rapid growth of TikTok social media has made it an effective digital promotion medium for Micro, Small, and Medium Enterprises (MSME). —The large number of comments left by users serves as a valuable source of information to measure customer perception and satisfaction. However, the characteristics of social media text, which are unstructured, informal, and contain slang and emojis, make manual analysis difficult. This study aims to classify customer review sentiments regarding MSME food products on TikTok into positive, negative, and neutral categories using the Multinomial Naive Bayes algorithm. The dataset consists of 1,041 clean comment entries collected via web scraping. Text preprocessing steps include case folding, slang normalization, stopword removal, and the removal of empty and duplicate data. Sentiment labeling was conducted using a lexicon-based approach, while feature weighting was performed using Term Frequency-Inverse Document Frequency (TF-IDF). The dataset was split into an 80% training set (832 data points) and a 20% testing set (209 data points). The model performance evaluation showed that the Naive Bayes algorithm achieved an overall accuracy rate of 75.60%. The model demonstrated high performance on the neutral class with an F1-score of 0.85 and a precision of 1.00 for the positive class. This research provides a practical contribution for MSME owners in evaluating customer feedback and developing more responsive digital marketing strategies‖.

Analysis Of Artificial Intelligence Deepfake Risks Regarding The Public's Ability To Distinguish Between Real And AI-Generated Content Using A Quantitative Descriptive Method

Muhammad Rizki Akbar, Pahrizal Pahrizal, Yulia Darmi, Rozali Toyib — Wed, 28 Oct 2026 02:11:06 +0700

The development of Artificial Intelligence (AI) technology has given rise to Deepfake technology capable of producing visual content resembling reality, creating a risk of disinformation and making it difficult for the public to distinguish real from AI-generated content. This study aims to analyze the public's ability to distinguish such content and the relationship between the public's knowledge of Deepfake AI (X1) and their ability to detect Deepfake content (Y1). This study uses a quantitative descriptive method with 100 social media user respondents aged 17-30 years, collected online via Google Forms from June 24 to July 24, 2026. The instrument consisted of five Likert-scale statements measuring X1 and ten pairs of real-Deepfake photos measuring Y1. Results show that detection ability is high (average 77.1%) and knowledge level is also high (average 3.87 on a 1-5 scale). Both instruments were valid and reliable (Cronbach's Alpha for X1 = 0.834; KR-20 for Y1 = 0.712). A Pearson correlation test showed no significant relationship between X1 and Y1 (r = -0.028; p = 0.784 > 0.05), confirmed by simple linear regression (R Square = 0.001), so the null hypothesis (H0) is accepted; conceptual knowledge was shown not to automatically improve actual detection ability. This study recommends improving digital literacy, AI content watermarking regulations, and refining the detection instrument in future research.

Decision Support System for Assessing Service Quality at Sawah Lebar Community Health Center Using the Weighted Product (WP) Method

Reno Supardi — Wed, 28 Oct 2026 02:11:27 +0700

The quality of healthcare services is a crucial factor in enhancing public satisfaction with the services provided by Community Health Centers (Puskesmas). Puskesmas Sawah Lebar requires a system to assist in assessing service quality in an objective, measurable, and systematic manner. This study aims to develop a Decision Support System (DSS) for evaluating service quality at Puskesmas Sawah Lebar by applying the Weighted Product (WP) method. The WP method facilitates decision-making based on predetermined criteria, enabling the generation of service quality scores and rankings. The criteria employed in this study encompass aspects such as service timeliness, staff friendliness, facilities, responsiveness, and information clarity. Research data were obtained through [questionnaires/interviews/observations] involving service users at Puskesmas Sawah Lebar. Each criterion was assigned a weight based on its level of importance, followed by a calculation process using the Weighted Product method to derive preference values for each alternative. The results indicate that the developed system assists the health center in conducting service quality assessments more rapidly, objectively, and in a structured manner. With this decision support system in place, the assessment results can serve as a basis for evaluation, helping Puskesmas Sawah Lebar improve the quality of services provided to the community.

Implementation Of Marker-Based Tracking And Fisher-Yates Shuffle Algorithms In An Augmented Reality-Based Solid Geometry Learning Application Using Unity For Mobile Platform

Ryandika Anugrah, Edy Hermansyah, Rizka Tri Alinse — Wed, 28 Oct 2026 02:11:52 +0700

Teaching three-dimensional shapes still faces challenges in visualizing three-dimensional objects, causing students to struggle with understanding abstract concepts. This study aims to design and develop an Augmented Reality (AR)-based geometric shapes learning application using Unity on the Android platform by implementing the Marker-Based Tracking and Fisher-Yates Shuffle algorithms. Marker-Based Tracking is used to detect markers in real-time and display three-dimensional geometric shapes, while the Fisher-Yates Shuffle algorithm is applied to randomize the order of questions in the assessment feature. This study employed the Research and Development (R&D) method with the Waterfall development model, which includes the stages of requirements analysis, design, implementation, testing, and maintenance. Data collection was conducted through observation, interviews, and a literature review. Application testing was conducted using Black Box Testing to evaluate system functionality and the System Usability Scale (SUS) to measure the application’s usability. The results show that the application was successfully developed using Unity and Vuforia on the Android platform. The implementation of Marker-Based Tracking successfully detected markers and displayed 3D geometric objects in real time, while the Fisher-Yates Shuffle successfully generated variations in the question order within the quiz feature. The results of the black-box testing showed that all of the application’s main features functioned 100% in accordance with the test scenarios. Meanwhile, the results of the SUS test administered to 20 students yielded an average score of 78.0, which falls into the “Acceptable” category with a Grade B (Good) rating. Thus, the developed application has a good level of usability and can be used as an interactive learning tool to help students understand three-dimensional geometry concepts.

Mobile-Based Drug Inventory Management And Prediction System Using Multiple Linear Regression (Case Study: Idola Bunda Pharmacy)

Rinaufal Liandisa, Yupianti Yupianti, Prahasti Prahasti — Wed, 28 Oct 2026 02:12:16 +0700

Manual drug inventory recording can delay recapitulation, increase recording errors, and make it difficult to estimate stock requirements for the next period. This study aimed to develop a mobile-based drug inventory management and prediction system for Idola Bunda Pharmacy and to apply Multiple Linear Regression to estimate inventory from historical data. A Research and Development design with a quantitative approach was employed using the Rapid Application Development model. Data were collected through observation, interviews, literature review, and transaction records covering purchases, sales, and inventory from January to December 2025. The system was implemented with PHP, MySQL, and Progressive Web Application technology. Sales and purchases were treated as independent variables, while inventory was the dependent variable. Testing with Woods Peppermint Syrup produced a January 2026 inventory prediction of 22 bottles compared with an actual value of 26 bottles. The Mean Absolute Percentage Error was 15.38%, which falls within the good prediction category, and black-box testing indicated that the main system functions operated according to the expected scenarios. These findings show that integrating transaction management and predictive analysis in a mobile-accessible application can support more structured inventory control and provide practical information for procurement decisions intended to reduce overstock and stockout risks.

Automated Data Extraction From Student Id Cards (KTM) Using Optical Character Recognition (OCR) And Regular Expressions Based On TKINTER GUI

Andes Putra Gunawan, Yulia Darmi, Ardi Wijaya, Yuza Reswan — Wed, 28 Oct 2026 02:12:36 +0700

This research aims to design and implement an automatic data extraction system from student ID card (KTM) images using a combination of Tesseract-based Optical Character Recognition (OCR) and Regular Expression, with a graphical user interface built using Tkinter. The developed system consists of an image preprocessing stage (text-zone cropping, skew correction with protection against textured backgrounds, and Otsu binarization), OCR text extraction using a multi-source and multi-page-segmentation-mode strategy, and data field extraction using tiered Regular Expression patterns supported by a keyword-mapping dictionary for the Study Program and Faculty fields. Testing was conducted on 100 student ID card images from Universitas Muhammadiyah Bengkulu using confusion matrix evaluation metrics, namely Precision, Recall, F1-Score, and Accuracy for each data field. The test results show that the system achieved an overall accuracy of 93.8%, with a Precision of 94.86%, Recall of 98.72%, and F1-Score of 96.75%. Per-field accuracy results show that the Student ID Number (NPM) field reached 100%, Faculty 98%, Study Program 95%, Place/Date of Birth 94%, and Name 82%. This research also identified and resolved a cross-platform issue in the OCR library that caused the loss of spaces and punctuation marks in the extraction results when run on the Windows operating system. The findings indicate that the combination of OCR and Regular Expression, specifically designed to match the visual characteristics of student ID cards, is able to produce an accurate and reliable data extraction system suitable for supporting academic administration processes.

Predicting Daily Patient Visits at Community Health Centers Using Machine Learning Algorithms to Optimize Medical Staff Scheduling

Zahza Ivana Putri, Edy Hermansyah, Dimas Aulia Trianggana — Wed, 28 Oct 2026 02:13:17 +0700

Optimal health care at primary-level health facilities depends heavily on the readiness of human resources. At the Lempuing Community Health Center in Bengkulu, there is a significant operational imbalance due to unpredictable fluctuations in the number of daily patient visits. Medical staff scheduling has thus far been conducted conventionally without considering future trends in patient arrivals. This results in long lines on certain days, while also causing labor inefficiency during periods of low patient volume. This study employs the Cross-Industry Standard Process for Data Mining (CRISP-DM) methodology. The modeling approach utilizes the Random Forest algorithm. The computational process leverages historical data on daily patient visits over the past three months, specifically from October through December 2025. In addition to building a computational prediction model, this study also integrates the forecasting results with the Standard Workload (SBK) formula. Based on the results of the Random Forest model testing, the evaluation performance showed a Mean Absolute Error (MAE) of 7.27 patients per day and a Root Mean Squared Error (RMSE) of 10.25 patients. The coefficient of determination obtained was 37.10%, which reasonably indicates the level of complexity and the highly stochastic nature of historical health data, which is susceptible to random external factors. Feature importance analysis revealed that the H-7 and H-1 Visit variables are the most dominant parameters influencing spikes in patient numbers. The system is effectively capable of transforming digital prediction figures into recommendations for medical staff allocation, ranging from 1 to 3 per day.

Design and Development of a Web-Based Bouquet Sales Information System with Transaction Management and Sales Reporting in Palembang City

Sagita Dwi Pradita, Dedi Haryanto, Jimmie Jimmie — Wed, 28 Oct 2026 02:20:11 +0700

The development of information technology encourages small and medium enterprises to digitalize their sales processes to improve transaction management and reporting efficiency. Bucketbungakita.plg, a bouquet sales business in Palembang City, still manages sales records and reports conventionally, causing difficulties in monitoring orders, payments, and sales information. This study aims to design and develop a web-based bouquet sales information system that supports transaction management and sales reporting. The system development method used is the System Development Life Cycle (SDLC) with the Waterfall model, consisting of requirement analysis, system design, implementation, testing, and evaluation stages. Customers can view products, manage shopping carts, place orders, complete checkout processes, make payments through bank transfers, and upload payment proofs. Administrators can manage products, transactions, payments, deliveries, and sales reports. Black box testing showed that all tested functions worked successfully according to the defined requirements. The developed system can support more structured transaction recording and provide easier management of sales reports.