Management of Information System Journal

Analisis Pemanfaatan Website Inaportnet Dengan Menggunakan Pendekatan Fuzzy Service Quality

Adhelia Salsabila, Amelia Yusnita, Rizky Zakariyya Rasyad — Fri, 10 Jul 2026 23:12:41 +0000

This study aims to analyze the effectiveness of the use of the Inaportnet website at PT BTJ Tanah Grogot to improve service quality and public trust. By using the Fuzzy Service Quality (Fuzzy SERVQUAL) approach, this study assesses the level of gap that occurs between user expectations and their perceptions in the field. through five dimensions of service quality. The results of data processing There are differences in each variable, which indicates that the discrepancy is negative, which indicates that the quality of service available has not reached the maximum point of meeting user expectations. The Reliabilitiy dimension is the main priority for improvement with the largest gap value of -0.85, while the Tangibles dimension has the best performance or is closest to user expectations with the smallest gap value of -0.20. Overall, managers are advised to make tactical improvements, especially in the aspect of system reliabilitiy for the smooth operation of maritime logistics.

Analisis Kepuasan Pengguna Aplikasi Mobile Banking Bankaltimtara Menggunakan Metode SERVQUAL

Aulia Fathul Jannah, Amelia Yusnita , Andi Yusika Rangan — Fri, 10 Jul 2026 23:27:04 +0000

Kemajuan teknologi perbankan digital mendorong peningkatan pemanfaatan aplikasi mobile banking sebagai alat transaksi
yang lebih praktis dan efisien. Meskipun demikian, masih terdapat selisih antara ekspektasi dan pengalaman pengguna mengenai
kualitas layanan aplikasi DG (Digital Generation) by Bankaltimtara. Studi ini dimaksudkan untuk mengukur tingkat kepuasan
pengguna serta Perbedaan kualitas layanan dianalisis menggunakan metode SERVQUAL. Data penelitian ini diperoleh dari hasil
pengisian kuesioner oleh 100 responden yang dianggap sebagai pengguna aktif aplikasi, dengan pengukuran berdasarkan lima dimensi,
yaitu Tangibles, Reliability, Responsiveness, Assurance, dan Empathy. Analisis dilakukan dengan menghitung selisih (gap) antara
persepsi dan harapan pengguna pada setiap dimensi. Hasil analisis menunjukkan bahwa skor kesenjangan di setiap dimensi bernilai
negatif dengan rata-rata sebesar -0,58, yang menandakan bahwa kualitas layanan belum mampu memenuhi harapan pengguna. Dimensi
Responsiveness memiliki nilai gap terbesar yaitu -0,76, sementara dimensi Assurance menunjukkan gap terkecil sebesar -0,52. Hal ini
mengindikasikan bahwa aspek daya tanggap menjadi prioritas utama yang perlu ditingkatkan. Oleh karena itu, langkah-langkah untuk
meningkatkan kualitas layanan sangat diperlukan secara menyeluruh untuk meningkatkan kepuasan pengguna serta mendukung
optimalisasi layanan digital perbankan.

Perancangan Chatbot Berbasis Flowise AI Untuk Layanan Informasi Pada Website Organisasi Kemahasiswaan

Muhammad Hudzaifiy Nur, Pitrasacha Adytia, Kusnandar — Fri, 10 Jul 2026 00:00:00 +0000

The delivery of information on the website of the HIMAPROSI student organization at STMIK Widya Cipta Dharma Samarinda is still one-way, requiring users to open each page individually to find the information they need. This study aims to design and implement a Flowise AI-based chatbot directly integrated into the HIMAPROSI website as a more interactive and efficient information service solution. System development uses the Waterfall method adapted into four stages: data collection, design, implementation, and testing. The chatbot was built using seven components on the Flowise AI platform: PDF File, Character Text Splitter, Jina Embeddings, In-Memory Vector Store, GroqChat (Llama 3.1 8b Instant), Conversational Retrieval QA Chain, and Buffer Memory. This ensures that every response generated is sourced directly from HIMAPROSI’s official documents. Testing was conducted using the Beta Testing method, involving 7 respondents who completed a questionnaire based on a Likert scale. The test results showed an average score of 84%, with the highest score of 89% in the aspects of ease of finding information and speed of obtaining information, proving that the developed chatbot is capable of providing accurate, interactive, and efficient information services for HIMAPROSI website users.

Implementasi Bi-LSTM Untuk Klasifikasi Sembilan Kategori Teks Berita Bahasa Indonesia

Ali Imron, Erna Daniati, Dwi Harini — Fri, 10 Jul 2026 23:52:52 +0000

This study aims to develop a classification model for Indonesian news texts across nine categories using the Bidirectional Long Short-Term Memory (Bi-LSTM) method. Unlike previous studies that predominantly utilized balanced datasets or focused on binary classification, this research evaluates the model's robustness under extreme real-world class imbalance, where the disparity in data volume between classes reaches over thirteen-fold. The research methodology encompasses text preprocessing (case folding, punctuation removal, and Sastrawi stemming), tokenization, padding, and dataset splitting at an 80:20 ratio. The model architecture integrates an embedding layer, a Bi-LSTM layer, a dropout layer, and a dense layer, which are optimized using a cost-sensitive class weight technique and the Adam optimizer. Experimental results demonstrate that the model achieved an accuracy of 0.85, with weighted average precision, recall, and F1-score values all at 0.85. The primary advantage of this method lies in its bidirectional feature extraction capability (forward and backward), which proves reliable in mitigating majority class bias and reducing semantic ambiguity within minority categories. The scientific contribution of this study provides empirical evidence and a comprehensive reference regarding the effectiveness of bidirectional contextual understanding in maintaining the performance stability of multi-class text classification within non-ideal, real-world data landscapes.

Proyeksi Tren Kategori Pakaian Mendatang Menggunakan Random Forest pada Data Transaksi Pelanggan

Nur Aini Umar, Andi Ircham Hidayat Hidayat, Eka Wijaya Paula — Sun, 12 Jul 2026 22:44:04 +0000

The dynamic fashion industry requires accurate trend projections for marketing and product development. This study aims to project future clothing trends using Random Forest as the primary model and XGBoost as the secondary model. The main dataset contains 3,900 transactions with demographic information, purchase history, seasonal data, and product categories. For local context, inventory data from the “Coffer Ruh” fashion store was integrated as a companion case study. The methodology included preprocessing, handling class imbalance with SMOTE, stratified splitting (80:20), training Random Forest and XGBoost, and evaluation using accuracy, precision, recall, F1-score, and a confusion matrix. Evaluation results (Outerwear, Footwear, Bottoms, Tops, Accessories) show that Random Forest achieved an accuracy of 67.56%, weighted precision of 68.04%, recall of 67.56%, and an F1-score of 67.79%, while XGBoost demonstrated similar performance with an accuracy of approximately 68%. The Random Forest model projected Jackets (17%), Coats (12%), and Shoes (10%) as the top three global trend categories. Store data analysis revealed the highest stock levels for children’s masks (35 pcs), red cornersticks (33 pcs, coats), and drams (31 pcs, jackets). There is some alignment: two of the three products with the highest inventory are outerwear items that align with global trends; however, masks are not a predicted apparel category. Due to limitations in the store data (small sample size, lack of time/transaction dimensions), transfer learning or hybrid dataset approaches cannot yet be applied, which is identified as a limitation and a direction for future research.

Analisis Sentimen Ulasan Aplikasi Mobile Legends Berbasis Pelabelan IndoBERT dan SMOTE dengan Komparasi Algoritma Klasifikasi Naive Bayes dan SVM

Muhammad Fauzan Aditiya Mufid, Erna Daniati, Arie Nugroho — Sun, 12 Jul 2026 23:05:55 +0000

This study was conducted to evaluate user sentiment trends toward the Mobile Legends: Bang Bang game app through reviews published on the Google Play Store platform. This research applied IndoBERT-based automatic labeling to generate sentiment labels that better represent the textual meaning of the reviews. The class distribution imbalance resulting from the labeling process was addressed using the Synthetic Minority Oversampling Technique (SMOTE) during the model training phase. The performance of two classification algorithms—the Naive Bayes Classifier and the Support Vector Machine (SVM)—was compared through hyperparameter tuning using GridSearchCV, with the F1-Macro metric. The results show that the Naive Bayes Classifier model delivers the best performance with an accuracy of 87.55%, an F1-Score of 85.78%, and an F1-Score of 56.70%. However, the model still exhibits significant limitations in recognizing the neutral sentiment class (F1-Score 0.21), which was further analyzed and found to be caused by the very small proportion of the neutral class (2.3% of the total data) as well as the characteristic of neutral reviews, which tend to be requests or suggestions to developers rather than explicit statements. This study contributes to the development of a more representative sentiment analysis methodology through a combination of transformer-based labeling, data imbalance handling, and classification algorithm comparison.

Evaluasi Kesuksesan Sistem TokSort Pada UMKM Devline Store Menggunakan Model DeLone dan McLean

Aditya Arya Respati, Erna Danianti, Aidina Ristyawan — Mon, 13 Jul 2026 01:17:02 +0000

Manual order data management in small and medium enterprises may cause work delays, recording errors, and difficulties in monitoring order completion. This study evaluates the success of TokSort as a mobile-based information system for processing Comma Separated Values (CSV) order data at Devline Store using an adapted DeLone and McLean IS Success Model. A descriptive quantitative approach with an evaluative case study framing was applied. The respondents were 11 internal TokSort users directly involved in order data management, so total sampling was used. Data were collected through a Likert-scale questionnaire and observation, then analyzed using mean values, validity testing, reliability testing, and exploratory Spearman Rank correlation. The results show that all variables were classified as good: system quality 3.42, information quality 3.64, use 3.86, user satisfaction 3.59, and net benefits 3.73. Spearman Rank results indicate positive and significant relationships among the tested variables. These findings show that TokSort is considered successful in supporting order data processing, grouping, and monitoring based on internal users’ perceptions. However, system stability and the ability to reduce order processing errors still require improvement. This study contributes an empirical evaluation of an internal SME operational information system based on CSV file processing.

Sistem Pendukung Keputusan Untuk Pemilihan Kualitas Telur Bebek Di Kabupaten Nganjuk Menggunakan Metode SAW

Angga Pradipa Eko Widodo Pradipa, Muhammad Najibullah Muzaki, Erna Daniati — Mon, 13 Jul 2026 02:35:23 +0000

The determination of duck egg quality among farmers and salted egg SMEs in Nganjuk Regency is still carried out manually based on experience, resulting in subjective, inconsistent, and time-consuming assessments. This condition causes the egg grading process to be less optimal and may affect the quality of products marketed. This study aims to design and develop a web-based Decision Support System (DSS) using the Simple Additive Weighting (SAW) method to support a more objective, faster, and structured duck egg quality selection process. The system was developed using the Waterfall model, consisting of requirements analysis, system design, implementation, and testing stages. The assessment was based on four criteria: egg weight (35%), egg size (10%), storage time (25%), and eggshell color scale (30%). The system was implemented using Google Apps Script with Google Spreadsheet as the database. The implementation results show that the system is able to perform normalization, weighting, preference value calculation, and automatic ranking of duck egg quality. Blackbox Testing on seven main modules, namely login, alternative data, criteria data, weight setting, SAW calculation, ranking results, and report printing, showed that all modules functioned according to user requirements with a success rate of 100%. The results indicate that the SAW method can support duck egg quality assessment more objectively, consistently, and efficiently than manual assessment, thereby assisting decision-making for farmers and salted egg SMEs in Nganjuk Regency.

Prediksi Harga Rumah Berbasis Machine Learning dengan Explainable AI untuk Interpretabilitas Faktor Penentu

Hadijah, Wiwin Handoko, Rizty Maulida Badri — Mon, 13 Jul 2026 03:15:59 +0000

House prices are determined by numerous interrelated factors, making it essential to develop prediction methods that are not only accurate but also interpretable by property business practitioners, investors, and policymakers. This study aims to construct a house price prediction model using a machine learning approach integrated with Explainable Artificial Intelligence (XAI) to produce predictions that are more transparent and comprehensibly interpretable. The data used in this study were derived from real property listings, incorporating several key variables including building area, land area, number of bedrooms, number of bathrooms, and garage capacity. Four machine learning algorithms were evaluated and compared, namely Linear Regression, Random Forest, XGBoost, and Gradient Boosting. The performance of each model was assessed using multiple evaluation metrics, comprising Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), coefficient of determination (R²), and Mean Absolute Percentage Error (MAPE). Experimental results demonstrate that the Random Forest algorithm achieved the best performance, yielding an R² value of 0.7636, MAE of IDR 1.305 billion, RMSE of IDR 2.037 billion, and MAPE of 25.84%. The best-performing model was subsequently analyzed using SHapley Additive exPlanations (SHAP) to provide both global and local model interpretability, as well as Local Interpretable Model-agnostic Explanations (LIME) to explain individual predictions at the instance level. The analysis reveals that building area and land area are the most influential factors in determining house prices. The proposed approach demonstrates a measurable improvement in model transparency, rendering prediction outcomes more comprehensible and trustworthy for end users.

Analisis Kepuasan Masyarakat Terhadap Pelayanan E-Ktp di Kantor Kecamatan Kembang Janggut Menggunakan Metode Servqual

Nur Syariffah, Ivan Haristyawan, Rizky Zakariyya Rasyad — Mon, 13 Jul 2026 07:14:55 +0000

E-KTP service is a form of public service that must be able to meet public expectations. However, various obstacles remain, such as the lengthy service process, lack of officer responsiveness, and technical constraints that have the potential to reduce public satisfaction. This study aims to analyze the level of public satisfaction with E-KTP services at the Kembang Janggut District Office using the SERVQUAL method. The study used a quantitative approach involving 83 respondents selected using the Slovin formula. Data were obtained through distributing questionnaires based on five SERVQUAL dimensions, namely tangibles, reliability, responsiveness, assurance, and empathy, then analyzed using the calculation of the gap value between public perceptions and expectations. The results showed that all dimensions had a negative gap value, which means the quality of service has not fully met public expectations. The responsiveness dimension had the largest gap value of -0.22, making it a top priority in service improvement, while the empathy dimension had the smallest gap value of -0.13, indicating that the attitude and attention of officers have been rated the best by the public. This research provides a contribution in the form of identifying service dimensions that are priority for improvement so that it can be a basis for the Kembang Janggut District Office in developing strategies to improve the quality of E-KTP services and increase public satisfaction.

Deteksi Real-Time Hama dan Penyakit Jamur pada Daun Tanaman Menggunakan Deep Learning Berbasis YOLO

Reni Yunita, Egi Dio Bagus Sudewo, Bela Astuti — Mon, 13 Jul 2026 07:24:09 +0000

Serangan hama pertanian dapat menurunkan produksi tanaman secara signifikan dan menghambat kegiatan pertanian. Kesulitan mendeteksi hama secara manual sejak tahap awal membuat petani sering menggunakan pestisida secara berlebihan, yang dapat menyebabkan pencemaran lingkungan dan risiko terhadap kesehatan. Untuk mengatasi permasalahan tersebut, berbagai sistem telah dikembangkan untuk membantu mendeteksi hama sejak dini sehingga petani dapat mengetahui lokasi hama secara lebih tepat. Namun, proses identifikasi di lapangan yang masih dilakukan secara manual sering memerlukan waktu lama, tenaga yang besar, serta rentan terhadap kesalahan identifikasi. Selain itu, beberapa sistem yang telah dikembangkan masih memiliki keterbatasan, seperti belum mampu melakukan deteksi secara real-time dan belum dilengkapi dengan sistem pemantauan berbasis web. Penelitian ini mengusulkan sistem deteksi hama dan penyakit jamur pada daun tanaman secara real-time menggunakan pendekatan deep learning berbasis arsitektur YOLO26. Model YOLO26 dipilih karena memiliki kemampuan deteksi objek yang cepat dan efisien sehingga cocok untuk aplikasi pemantauan pertanian secara real-time. Dataset yang digunakan terdiri dari citra daun tanaman yang telah dianotasi ke dalam dua kelas objek, yaitu pest dan fungus. Hasil pengujian menunjukkan bahwa model yang diusulkan mampu mencapai precision sebesar 82%, recall sebesar 71%, dan mAP@0.5 sebesar 78%, dengan nilai mAP@0.5–0.95 sebesar 37,4%. Selain itu, model memiliki waktu inferensi sekitar 12,1 ms/citra, sehingga mampu melakukan deteksi secara real-time. Secara keseluruhan, sistem yang dikembangkan berpotensi membantu petani dalam melakukan pemantauan tanaman secara otomatis dan mendeteksi serangan hama serta penyakit jamur sejak dini, sehingga dapat mengurangi kerusakan tanaman, menekan penggunaan pestisida secara berlebihan.

Sistem Deteksi Jatuh Lansia Real-Time Berbasis YOLOv8 dengan Notifikasi Telegram dan Dashboard Web

Claudio Syanu Mareta Dinata, Rina Firliana, Arie Nugroho — Mon, 13 Jul 2026 09:46:33 +0000

Falls in the elderly represent a serious global health problem. According to the WHO, one in three elderly people aged over 65 years experiences a fall each year. In Panti Werda Kediri, monitoring of elderly activities is still performed manually by staff, causing fall incidents to go undetected quickly. Previous YOLO-based fall detection studies generally produce models without integrating them into monitoring platforms usable by non-technical end users and without automatic notification. This research aims to determine the effectiveness of a monitoring system in detecting normal activities and fall incidents in elderly residents in real time at Panti Werda Kediri using YOLOv8. The system was developed using the Waterfall method through stages of requirements analysis, system design, implementation, and testing. The detection component uses a retrained YOLOv8 model to recognize two classes: normal and fall. The backend is built with FastAPI and PostgreSQL, equipped with a web-based monitoring dashboard and automatic notifications via Telegram Bot. A fall confirmation mechanism based on 3 consecutive frames with a 1.5-second cooldown suppresses false positives. Blackbox testing conducted at Panti Werda Kediri shows all 10 test scenarios passed. The system successfully sends real-time Telegram notifications in under 2 seconds with visual evidence each time a fall is confirmed, provides live camera streaming, and displays complete detection history through a web dashboard accessible to non-technical staff.

Emergency Report System Multi-Channel Notification Berbasis Web pada Dinas Damkar Kabupaten Kediri

Erlangga Fajar Ramadhan Ramadhan; Rini Indriati, Dwi Harini — Tue, 14 Jul 2026 22:54:42 +0000

Conventional emergency reporting at the Kediri Regency Fire Department suffers from three structural problems: location inaccuracy, the absence of self-tracking for the public, and unstructured documentation. Similar studies typically address these issues separately, without integrating automatic geolocation, multi-channel notification, role-based access control (RBAC), and security validation into one empirically tested system. This study develops a web-based Emergency Response Information System (SIRTADA) using the Waterfall method, with a three-tier architecture combining automatic geolocation, Firebase Cloud Messaging, and a WhatsApp Gateway as parallel notification channels. Multi-level evaluation shows solid results: 100% unit test pass rate (63 test cases), 97.5% integration test pass rate (78/80 assertions), and an 89.3% UAT score ("Very Good"), although a Stored XSS vulnerability requiring further mitigation was identified. The novelty of this study lies in integrating automatic geolocation, multi-channel notification, and RBAC into a single empirically validated web-based emergency response architecture, offering an evaluation framework adoptable for similar public service systems

Pengembangan Aplikasi Helpdesk IT Berbasis Web dengan Integrasi Inventaris dan Logbook di RSUD Gambiran

Moh. Teguh Purwanto, Rini Indriati, Arie Nugroho — Tue, 14 Jul 2026 23:03:41 +0000

At RSUD Gambiran Kediri City, the coordination of IT issue resolution has long relied on informal communication via telephone calls and WhatsApp messages, resulting in incomplete incident reports, difficulties in tracking resolution status, and a lack of service documentation trails. This condition risks directly hindering the quality of healthcare services, as disruptions to information technology infrastructure affect all hospital work units. To address these issues, this study developed a web-based IT helpdesk information system designed to integrate directly with the hospital’s existing inventory and logbook applications. The development process adopted the Waterfall model through four sequential phases: requirements analysis, design, implementation, and testing. Requirements data were gathered through interviews, field observations, and questionnaires distributed to IT officers and relevant department heads. The technology stack utilizes PHP Laravel as the primary framework, MySQL for database management, and Bootstrap for the user interface, hosted on an Ubuntu server running inside an LXC container on the Proxmox virtualization platform. Key features include a six-stage tiered complaint handling workflow, real-time notifications via Telegram, and API connections to the inventory and logbook applications. Functional verification was conducted using Black Box Testing across 14 manual scenarios, 63 feature tests, and 4 unit tests, all of which passed. A feasibility assessment by three respondents through User Acceptance Testing demonstrated full acceptance of all tested features. A satisfaction questionnaire distributed to ten IT officers yielded an average score of 4.6 out of 5. The resulting system successfully replaced previous informal reporting patterns while enhancing organization, response time, and information transparency in managing hospital IT complaints

Analisis Sentimen Ulasan EA FC 26 Di Steam: Perbandingan SVM Dan Multinomial Naïve Bayes

Rafi Adha Azhar, Wahyu Widodo — Tue, 14 Jul 2026 23:09:01 +0000

User reviews on the Steam platform contain important information about player perceptions of EA FC 26. However, the large volume of reviews and their unstructured textual form make manual analysis difficult and prone to inconsistency. This study aims to identify user sentiment tendencies and compare the performance of Support Vector Machine (SVM) and Multinomial Naïve Bayes in classifying EA FC 26 reviews on Steam. The research follows the Knowledge Discovery in Database (KDD) stages, consisting of selection, preprocessing, transformation, data mining, and evaluation. A total of 5,000 recent English-language reviews were collected through the Steam Review API. After data processing, 4,988 valid reviews were obtained, consisting of 2,823 negative reviews and 2,165 positive reviews based on the voted_up recommendation status. The review texts were represented using TF-IDF N-Gram features consisting of unigrams and bigrams, then classified using SVM and Multinomial Naïve Bayes. The evaluation results show that SVM achieved an accuracy of 82.26% and a weighted F1-score of 0.8230, while Multinomial Naïve Bayes achieved an accuracy of 77.15% and a weighted F1-score of 0.7598. The McNemar test produced a p-value of 0.000421, indicating a statistically significant difference between the two models. This study contributes by providing a comparative evaluation based not only on general accuracy, but also on per-class performance, confusion matrix, prediction error patterns, and statistical significance.

Media Virtual Reality Berbasis Pemodelan 3D untuk Edukasi Warisan Budaya Paseban Tri Panca Tunggal

Yulyanto, Fachri El'Kariem, Rika Nugraha — Tue, 14 Jul 2026 23:25:38 +0000

The erosion of cultural heritage literacy among younger generations is a pressing challenge in Indonesia's digitizing society. Paseban Tri Panca Tunggal, a Sundanese heritage building constructed in 1840 in Cigugur, Kuningan, West Java, embodies significant philosophical and architectural values of the Sunda Wiwitan belief system, yet remains largely inaccessible to students through conventional educational approaches. This study presents the design and implementation of a 3D modeling-based Virtual Reality (VR) educational medium for this heritage site using the Multimedia Development Life Cycle (MDLC) framework. Blender was used for hardsurface 3D modeling, Unity for interactive integration, and Spatial.io for metaverse-based VR deployment, producing an immersive learning experience accessible via web browser, Android devices, and VR headsets. Nine key architectural elements, each bearing distinct philosophical meaning, were digitally reconstructed with interactive audio narration and informational pop-ups. Expert media validation yielded 4.3/5.0 (Very Feasible), object accuracy validation scored 3.5/5.0 (Feasible), and user testing with 36 Grade-5 primary students produced 4.21/5.0 (84.2%, Very Feasible). These results demonstrate that VR-based 3D modeling is an effective medium for cultural heritage education, offering a replicable model for digital preservation of locally significant heritage sites in Indonesia. As the first VR educational medium specifically designed for Paseban Tri Panca Tunggal, this work contributes a replicable digital preservation framework for locally significant heritage sites across Indonesian educational contexts.

Game RPG Matematika Berbasis Fuzzy Mamdani-DDA dan Dasbor Guru untuk Sekolah Dasar

Yulyanto, Muhammad Nauval Bintang, Rio Andriyat Kridiawan — Fri, 17 Jul 2026 02:21:33 +0000

This study develops and evaluates the feasibility of “Jiwa Menghitung,” an Android turn-based RPG for elementary school mathematics (Grades 4–6). The principal novelties are: (1) adoption of Fuzzy Mamdani inference—rather than Sugeno—as the DDA engine, yielding linguistically interpretable categorical outputs (Easy/Medium/Hard) directly communicable to educators; and (2) a dual-sided architecture integrating the game client with a real-time teacher dashboard via Firebase. The Mamdani engine, governed by nine IF-THEN rules and Centroid of Gravity (COG) defuzzification, adapts question difficulty from two inputs: player response time and error count. Membership function parameters were validated through expert review by two mathematics teachers, and the rule base verified by white-box analysis (V(G)=3). User Acceptance Testing (UAT) with 42 students yielded 90.98% overall acceptance (Usability: 95.35%; DDA System: 80.00%), while teacher UAT (n=2) produced 88.09%—both exceeding the “Very Feasible” threshold. N-gain analysis for Grades 4 and 5 produced moderate gains (g=0.36 and g=0.34), while Grade 6 exhibited a ceiling effect identifying a critical design constraint. Findings confirm the technical feasibility and user acceptance of Mamdani-based DDA, while establishing adaptation granularity as a key development priority.

Optimasi Distribusi Makan Bergizi Gratis berbasis Vehicle Routing Problem with Time Windows (VRPTW)

Fri, 17 Jul 2026 03:24:05 +0000

The Free Nutritious Meal Program (MBG) requires an efficient distribution system to ensure the timely delivery of meals while maintaining food quality and safety. This study aims to implement the Vehicle Routing Problem with Time Windows (VRPTW) method in a web-based MBG Distribution Information System to optimize delivery routes based on vehicle capacity, demand quantity, destination locations, and service time window constraints. The research methodology consisted of problem identification, literature review, data collection, system design, implementation, and system evaluation. The study utilized operational data from the Nutrition Fulfillment Service Unit (SPPG) in Desa Baru Pasar VIII, Tanjung Beringin, Langkat Regency, serving 13 schools with a total demand of 1,672 meal portions using two delivery vehicles. The implementation results showed that the VRPTW method generated an optimized distribution route with a total travel distance of 42.74 km. The evaluation results demonstrated that VRPTW reduced the total travel distance from 49.86 km to 42.74 km, representing a 14.28% improvement in distribution efficiency. Furthermore, the total distribution time decreased from 118 minutes to 101 minutes, achieving a 14.41% reduction compared with the conventional distribution method. In addition, the system automated route planning, minimized route overlap, and improved vehicle capacity utilization. These findings indicate that integrating the VRPTW method into the MBG Distribution Information System enhances operational efficiency and supports more effective and timely meal distribution.

Analisis Kepuasan Mahasiswa Terhadap Layanan BAAK STMIK WICIDA Menggunakan Metode Servqual

Gia Lestari Geaa, Ivan Haristyawan, Rizky Zakariyya Rasyad — Fri, 17 Jul 2026 03:38:43 +0000

The Academic and Student Affairs Administration Office (BAAK) plays an important role in supporting academic services and influencing student satisfaction. However, there is still a gap betwìeìen thìe sìervicìes rìecìeivìed by studìents and thìeir ìexpìectations, making it nìecìessary to ìevaluatìe thìe quality of thìe sìervicìes providìed. This study aims to dìetìerminìe thìe lìevìel of studìent satisfaction with thìe sìervicìes of BAAK at STMIK WICIDA using thìe SERVQUAL mìethod. This rìesìearch ìemployìed a quantitativìe approach by distributing quìestionnairìes to studìents. Thìe collìectìed data wìerìe analyzìed by comparing studìents' pìercìeptions and ìexpìectations basìed on thìe fivìe SERVQUAL dimìensions, namìely tangiblìe, rìeliability, rìesponsivìenìess, assurancìe, and ìempathy. Thìe rìesults indicatìe that all sìervicìe dimìensions havìe nìegativìe GAP valuìes, mìeaning that thìe sìervicìes providìed by BAAK havìe not yìet fully mìet studìents' ìexpìectations. Thìe tangiblìe dimìension has thìe largìest GAP valuìe of -0.68, followìed by rìeliability (-0.23), rìesponsivìenìess (-0.16), assurancìe (-0.12), and ìempathy (-0.07). Thìesìe findings providìe an ovìerviìew of thìe sìervicìe aspìects that should bìe prioritizìed for improvìemìent, particularly thìe tangiblìe dimìension, and can sìervìe as ìevaluation matìerial for thìe institution in improving thìe quality of acadìemic and studìent sìervicìes.

Evaluasi Komparatif Forecasting PNBP BLU Berbasis SARIMAX terhadap Target Renstra

Roberto, Abdul Hamid Arribathi — Fri, 17 Jul 2026 03:47:09 +0000

This study evaluates forecasting methods for predicting BLU X non-tax state revenue and examining its alignment with the 2025-2029 strategic plan targets. The dataset consists of monthly PNBP transactions from January 2021 to April 2026 aggregated by month, Chart of Account (COA), and department. Four approaches are compared: Holt-Winters, SARIMAX, XGBoost, and bottom-up forecasting based on dominant COA groups. Model performance is assessed on the January 2025-April 2026 test period using Mean Absolute Error, Root Mean Squared Error, and Mean Absolute Percentage Error. The results indicate that SARIMAX(1,1,1)x(0,0,0,0) performs best with a MAPE of 17.97%. The selected model is then refitted using the complete historical dataset and applied to forecast revenue through December 2029. The comparison with strategic plan targets reveals a narrow gap in 2025-2026, while 2027-2029 projections are slightly above the target. These findings show that forecasting can support evidence-based budgeting, target realism assessment, and monitoring of dominant revenue services.

Dashboard Rekomendasi Bundling Produk Berbasis Algoritma Apriori Terintegrasi pada Sistem Point of Sale

Akbar Zakaria, Suparyanto — Fri, 17 Jul 2026 23:15:28 +0000

Competition in the minimarket retail business requires owners to implement data-driven promotional strategies rather than relying on intuition. Toko Sahabat, the case study of this research, holds a large volume of sales transaction data that has not been optimally utilized to support business decision-making, making it difficult for the owner to determine which products should be promoted or bundled together. This research aims to implement the Apriori algorithm into a product bundling recommendation dashboard integrated within a web-based Point of Sale (POS) application. The system was built using the Laravel 11 framework and MySQL 8.0 database, using 14,166 transactions collected from December 1, 2025, to June 20, 2026. The Apriori algorithm was used to identify frequent itemsets and generate association rules based on support, confidence, and lift ratio values. Using a minimum support of 2% and minimum confidence of 30%, the system generated 17 valid association rules, all with lift ratio values greater than one, the strongest being between Gula Pasir Gulaku and Kopi Kapal Api (confidence 90.27%, lift 5.42), which were mapped into five ready-to-use bundling package recommendations for shelf arrangement and promotional strategies. Black box testing across eight scenarios achieved a 100% success rate, confirming that all system functions operate as designed and are ready for operational use. The scientific contribution of this research lies in the direct integration of the Apriori algorithm into an active POS system, enabling bundling recommendations to be generated automatically and continuously from real transaction data, in contrast to prior studies that typically present association analysis results manually and separately from the store's operational system.

Analisis Dampak ChatGPT Sebagai Code Assistant Terhadap Kualitas Kode Mahasiswa Informatika UTS

Siska Atmawan Oktavia Siska, Siska Atmawan Oktavia, Muhammad Alfan Habib — Mon, 20 Jul 2026 00:00:00 +0000

The rapid advancement of Artificial Intelligence (AI) has introduced various tools that support software development, one of which is ChatGPT. This study aims to analyze the impact of using ChatGPT as a code assistant on the quality of program code produced by Informatics students at Universitas Teknologi Sumbawa. A quantitative experimental method was employed by comparing programming tasks completed manually and with the assistance of ChatGPT. Code quality was evaluated using SonarQube based on three metrics: Maintainability Index, Cyclomatic Complexity, and Reliability (Bug Count), followed by statistical analysis to examine differences between the two conditions. The results indicate that there were no significant differences across all evaluated metrics between manually written code and code generated with ChatGPT assistance (p > 0.05). These findings suggest that the use of ChatGPT did not affect code quality in this study; however, it still has the potential to improve the efficiency of software development. Furthermore, this study provides empirical evidence regarding the impact of using ChatGPT on code quality, based on SonarQube static analysis metrics. The findings are expected to serve as a reference for educators, students, and researchers in evaluating the use of ChatGPT as a code assistant in both learning and software development contexts.

Penerapan Metode Linear Regression untuk Prediksi Pendapatan UMKM Kantin Sekolah pada Program Makan Bergizi Gratis

Yessica Siagian, Jeperson Hutahaean, Andri Nata, Hikmat Syahputra Tarigan — Mon, 20 Jul 2026 01:29:38 +0000

Penelitian ini bertujuan menganalisis dan memprediksi dampak Program Makan Bergizi Gratis (MBG) terhadap pendapatan UMKM kantin sekolah di Kabupaten Asahan menggunakan metode regresi linear sederhana. Penelitian menggunakan pendekatan kuantitatif dengan data yang diperoleh dari 28 pelaku UMKM kantin sekolah yang terdampak pelaksanaan Program MBG. Analisis dilakukan untuk mengetahui hubungan antara variabel independen dan perubahan pendapatan UMKM serta membangun model prediksi yang dapat digunakan sebagai dasar pengambilan keputusan. Hasil penelitian menghasilkan model regresi Y = 157.321 − 46,35X yang menunjukkan adanya hubungan antara variabel yang dianalisis. Hasil analisis juga menunjukkan bahwa pedagang makanan berat mengalami dampak penurunan pendapatan paling besar, sedangkan pedagang non-makanan pokok relatif lebih mampu mempertahankan bahkan meningkatkan pendapatannya. Evaluasi model menggunakan Mean Absolute Percentage Error (MAPE) menghasilkan nilai 12,45%, yang termasuk dalam kategori baik, sehingga model dinilai mampu memberikan prediksi dengan tingkat akurasi yang memadai. Hasil penelitian diharapkan dapat menjadi bahan pertimbangan bagi pemerintah dan pihak sekolah dalam menyusun kebijakan pendampingan bagi UMKM yang terdampak Program MBG.

Penerapan Sistem Lampu Jalan Automatis Berbasis Sensor Cahaya Dengan Integrasi IoT

Mon, 20 Jul 2026 01:56:05 +0000

The development of Internet of Things (IoT) technology provides significant opportunities in automation systems, including public street lighting systems. Manual operation of street lights often causes energy waste because the lights remain active when unnecessary. This research aims to design and implement an automatic street lighting system based on light sensors integrated with IoT using NodeMCU ESP8266 and Light Dependent Resistor (LDR) sensors. The system is designed to automatically control the lights based on environmental light intensity and provide real-time monitoring through the Blynk application. The research method used is an experimental method involving hardware design, programming, implementation, and testing stages. The results show that the system can accurately detect changes in light intensity and automatically control the street lights according to environmental conditions. In addition, monitoring data can be displayed in real-time through the Blynk application via internet connectivity. The system also improves energy efficiency because the lights only operate when needed. Based on testing results, the system works properly, stably, and has potential applications in smart lighting and smart city infrastructure.

Analisis Business Intelligence Ulasan Negatif Aplikasi M-Pajak Menggunakan BERTopic untuk Evaluasi Layanan Digital

Henry Pandia — Wed, 22 Jul 2026 17:41:28 +0000

The M-Pajak application is part of the government’s digital tax service transformation aimed at improving the efficiency and accessibility of tax services for the public. However, user complaints on the Google Play Store indicate that technical and operational problems persist with the application. This study aims to analyze negative user reviews of the M-Pajak application using BERTopic to generate Business Intelligence insights related to the quality of government digital tax services. BERTopic was selected because it uses transformer-based contextual embeddings, which are well-suited to analyzing short, unstructured, and informal mobile application reviews, thereby helping identify complaint themes in a more context-aware way than traditional word-distribution-based topic modeling approaches. The data were collected from Google Play Store reviews from May 2023 to May 2026, totaling 6,024 user reviews. A total of 5,008 negative reviews with ratings of 1 or 2 were used as the primary focus of the analysis. The research stages included data scraping, preprocessing, BERTopic modeling, topic evaluation, and Business Intelligence interpretation. The results show that the main user problems are related to authentication and login, system errors, email and OTP verification, NPWP and NIK registration, and digital service stability. Model evaluation yielded a Topic Coherence score of 0.522, a Topic Diversity score of 0.667, and a Topic Quality score of 0.348, indicating that the topic quality was moderate yet interpretable. The contribution of this study lies in utilizing negative reviews of a digital tax service application as a source of operational insights to support the evaluation and improvement of government digital public services.

Sistem Pendukung Keputusan Pemilihan Guru Berprestasi Menggunakan Metode SAW

Bernadus Gunawan Sudarsono, Rakhmi Khalida — Wed, 22 Jul 2026 17:57:54 +0000

Sistem Pendukung Keputusan atau Decision Support System (DSS) adalah sebuah sistem berbasis komputer yang membantu pengambilan keputusan dalam situasi semi-terstruktur dan tidak terstruktur. Salah satu metode yang digunakan dalam SPK adalah Multi Criteria Decision Making (MCDM), dengan metode Simple Additive Weighting (SAW) sebagai salah satu teknik populernya. Penelitian ini bertujuan untuk membangun Sistem Pendukung Keputusan (SPK) guna menentukan guru berprestasi dengan menggunakan metode SAW. Langkah-langkah dalam metode SAW meliputi: menentukan kriteria dan bobotnya, menilai setiap guru berdasarkan kriteria yang telah ditetapkan, melakukan normalisasi nilai untuk setiap kriteria, menghitung skor total setiap guru dengan mengalikan nilai normalisasi dengan bobot kriteria, dan akhirnya membuat peringkat guru berdasarkan skor total yang diperoleh. Kriteria yang digunakan dapat mencakup kualitas pengajaran, inovasi dalam pembelajaran, dan kontribusi di luar kelas, dengan masing-masing kriteria diberi bobot sesuai tingkat kepentingannya. Hasil akhir dari sistem ini adalah daftar peringkat guru berprestasi yang mencerminkan prestasi mereka secara objektif, sehingga mempermudah dalam proses pemilihan guru berprestasi terbaik. Dengan demikian, SPK menggunakan metode SAW ini dapat meningkatkan transparansi dan akurasi dalam penilaian prestasi guru.

Perancangan Sistem Pendukung Keputusan Berbasis Rekayasa Perangkat Lunak Membantu Penentuan Prioritas Penanganan Perkara Hukum

Irianto, Abdul Karim Syahputra, Sumantri — Wed, 22 Jul 2026 18:21:25 +0000

The increasing number of legal cases requiring resolution has created significant challenges for law enforcement agencies in determining case handling priorities in an objective, timely, and consistent manner. This study aims to design a Software Engineering-based Decision Support System (DSS) to assist in determining the priority of legal case handling based on predefined evaluation criteria. The Simple Additive Weighting (SAW) method is employed to perform the weighting and ranking of case alternatives, while the software development process follows the Waterfall model, consisting of requirements analysis, system design, implementation, testing, and maintenance. The evaluation criteria include case urgency, social impact, number of victims, case complexity, and case resolution deadlines. System evaluation is conducted through functional testing using Black Box Testing, software quality assessment based on the ISO/IEC 25010 quality characteristics (functional suitability, usability, and performance efficiency), and recommendation accuracy measurement by comparing the system-generated rankings with expert judgments using the Spearman Rank Correlation coefficient. The expected system performance targets include 100% functional success, a minimum usability score of 85% (excellent category), a system response time of less than 3 seconds, and a ranking correlation coefficient of ≥0.90, indicating a very high level of agreement between the system recommendations and expert decisions. The expected outcome of this research is a Decision Support System capable of improving objectivity, consistency, and efficiency in prioritizing legal case handling while reducing subjectivity in the decision-making process.

Rancang Bangun Sistem Monitoring Suhu Pasien Rawat Inap Berbasis Internet of Things

Dedi Suarna — Sat, 25 Jul 2026 12:36:20 +0000

Monitoring the health condition of inpatients is an important aspect of hospital services to ensure optimal patient monitoring. The manual process of monitoring body temperature has several obstacles, such as limited time for medical personnel, delays in obtaining information on patient conditions, and the ineffectiveness of the continuous monitoring process. This study aims to design and build an Internet of Things (IoT)-based inpatient body temperature monitoring system that is capable of providing real-time patient condition information. This system uses a DS18B20 temperature sensor as a patient temperature reader and a NodeMCU ESP8266 as a microcontroller that functions to process and transmit data via the internet. The temperature data obtained is displayed on the Blynk application on a smartphone so that it can be monitored by medical personnel anytime and anywhere. The research methods used include observation, interviews, literature studies, system design, hardware and software implementation, and system testing. The results of the study indicate that the system built is able to read patient body temperature well, send data to the Blynk server, and display temperature information in real-time on a smartphone device. Thus, this IoT-based patient health monitoring system can assist medical personnel in monitoring patient conditions more effectively, quickly, and efficiently.

Machine Learning-Based Predictive System for Cultural Heritage Site Condition Assessment

Arif Ridho Lubis, Ali Basrah Pulungan — Sat, 25 Jul 2026 12:44:14 +0000

Cultural heritage preservation in North Sumatra faces challenges from manual, decentralized, and reactive site condition reporting. This study designs and develops a web-based Cultural Heritage Information System (CHIS) integrated with a Machine Learning (ML) prediction module for the Disbudparekraf of North Sumatra Province. The system is built using the CodeIgniter 4 framework with Role-Based Access Control (RBAC), MariaDB database, GIS integration via Leaflet.js, and a prediction module employing Random Forest Regressor with features including health score, structural integrity, physical integrity, authenticity, age factor, and maintenance score. Development follows the Waterfall model encompassing requirements analysis, design, implementation, black-box testing (47 test cases), and deployment. Results demonstrate successful integration of 12 historical assessment records from 10 priority heritage sites, generating 8 predictions with an average confidence score of 45.9% and automatic identification of high-risk sites such as Masjid Raya Al Mashun (confidence 51%). The system produces structured maintenance recommendations across three priority categories with specific timelines. The moderate confidence score reflects initial dataset limitations and is expected to improve with accumulating assessment data. This research contributes a replicable GIS-ML integration model for heritage conservation at the Indonesian local government level.

Pengembangan Aplikasi SPMI Berbasis Web Untuk Monitoring Tri Dharma Dan Audit Mutu Internal Di Universitas Cipasung Tasikmalaya

Sat, 25 Jul 2026 12:57:39 +0000

The monitoring process of the Tri Dharma Perguruan Tinggi (Three Pillars of Higher Education), learning monitoring, and Internal Quality Audit (AMI) at Cipasung University, Tasikmalaya, is still carried out manually, so data management and quality evaluation processes have not been running optimally. This study aims to develop a web-based Internal Quality Assurance System (SPMI) application that can support the monitoring and quality management process in a more integrated manner. The research method used is the Prototype method with the stages of needs analysis, system design, prototype development, and system testing. The application is designed using the Laravel framework and MySQL database. This study also applies RabbitMQ as a message broker to support asynchronous data communication between system modules. The main features of the application include Semester Learning Plan (RPS) management, research monitoring, Community Service (PKM), Intellectual Property Rights (IPR), and Internal Quality Audit (AMI) management. Testing results using the blackbox testing method indicate that all main features of the system can run according to user needs. The novelty of this study lies in the application of RabbitMQ in the web-based SPMI system to support the monitoring process and information distribution more quickly and integratedly.