Journal of Artificial Intelligence and Engineering Applications (JAIEA)

Design and Construction of a Website-Based Posyandu Service System in Sukaraja Village

Putri Saniyyah — 2026-06-15

Posyandu services at the auxiliary health center are a place to carry out health service activities. The auxiliary health center (pustu) still carries out posyandu service activities for the community. Posyandu services are located in Sukaraja Village, South Prabumulih. Currently, health services in Sukaraja Village are still providing services manually and do not have a website. The purpose of this study is to build a website so that officers and the community can more easily record identity data. Based on the results of the study, the researcher tried to build a web-based service system using a qualitative descriptive method by collecting data in the form of observations, interviews, literature studies and developed using the Rapid Application Development (RAD) method. System design tools use Use Case, Activity Diagram and Class Diagram.

Website Design for the Sukarami Village Head's Office

Femi Mutiara — 2026-06-15

The Sukarami village head office is under the auspice of the village government which has an Organization structure the sukarami village head office is the village head office located in sukarami village,rambang district, muara enim regency, south Sumatra province.Currently, the sukarami village head office does not have a website so that everyone can find out about the village head office to find out information such as village profiles, history of village activities, village programs,village structures, infrastructure owned by the village, and information on social assistance from the government and regions.The purpose of this study is to build a website so that the community can more easily find out about what is in the sukarami village head office. The research method uses a descriptive qualitative method with data collection in the form of observation,interviews and literature studies.The type of data consists of qualitative data, and data sources consist of primary data and secondary data.The device development method uses the rapid application development method.The system design tools used are the care diagram class diagram the website uses the PHP and MySQL programming languages.

Implementation of the Rapid Application Development (RAD) Model in the Web-Based MSME Product Sales System

Riski Dwi Anjani — 2026-06-15

A web-based sales system is an effective solution to help micro, small, and medium enterprises (UMKM) manage sales efficiently and in a structured manner. This research was conducted at Anisa Store, one of the UMKM in Majasari Village, Prabumulih, which still uses a manual sales process, so it has difficulty in recording data. The purpose of this study is to build a web -based sales system using the Rapid Application Development (RAD) Model, which allows the system development process to be carried out quickly and actively involves users. The RAD model consists of several stages, namely needs planning, design, construction and relocation. The result of this study is a web-based sales system that can be used by the owner of Anisa Store to manage product data, sales transactions, and shipping. By implementing this system, it is expected to increase the marketing reach of Anisa Store.

Implementation of Village Management Information System for Monitoring and Evaluation of Pangkul Village Development Program

Zaiyanah Putriyani — 2026-06-15

Pangkul Village is optimizing monitoring and evaluation of development programs through the implementation of a new information system. This system was developed using agile methods to adapt to rapidly changing needs. By implementing agile methods in monitoring and evaluating village development programs, it is hoped that a more adaptive and results-oriented process can be created that is beneficial to the community. This activity aims to identify progress, challenges, and impacts of development programs, as well as ensure community participation in the village development process. The evaluation results show the development of village government performance in planning and implementation, as well as the identification of relevant performance indicators for continuous improvement

Android-Based School Exam Information System for Sixth Grade Students at SD Negeri 23 Prabumulih

Myke Lastri Miyanti Myke — 2026-06-15

The school examination information system is a digital platform that aids in the administration of examinations for 6th-grade elementary school students. This research aims to develop an efficient and effective school examination information system to enhance the examination administration process at the elementary school level. The system development method waterfall and utilizes an object-oriented software design approach, considering the specific needs of elementary school examination administration. The developed system encompasses various features, such as examination scheduling, student data management, examination result recording, and provision of evaluation reports. The implementation of this system is expected to provide convenience for schools in managing and conducting school examinations, thereby facilitating smoother and more efficient learning processes.

Design and Development of a Web-Based Integrated Health Post Application for Infants, Toddlers, and Pregnant Women at the Belida Darat Community Health Center

Merlianda Oktarina — 2026-06-15

Belida Darat Health Center is under the auspices of the Muara Enim Regency Health Office, this Health Center is located in Belida Darat District, precisely in Tanjung Bunut Village. Currently, Belida Darat Health Center does not have an application for managing data on infants, toddlers and pregnant women, so the purpose of this study is to build a posyandu application to facilitate the management of data on infants, toddlers and pregnant women at Belida Darat Health Center. The research method uses a qualitative descriptive method with data collection techniques in the form of observation, interviews and literature studies, data sources consist of primary data and secondary data. while for the system development method using the RAD (Rapid Application Development) method, the system design tool used is UML (unified modeling Language). The design of this application uses the PHP (Processor Hypertext) programming language, MySQL Database and Coding using Visual Studio Code.

Website-Based Administrative Governance Information System in Lubuk Semantung Village

Suri Purnama Sari — 2026-06-15

Lubuk Semantung Village is optimizing administrative governance through the implementation of a new information system. This system was developed using agile methods to adapt to rapidly changing needs. With the integration of modules such as population data management, financial management, activity documentation, and reporting, this system aims to improve efficiency and transparency. Initial results from the implementation show a significant increase in the accessibility of information for residents and the accountability of the village government. This information system is expected to be a pilot model for other villages in improving the quality of administrative governance.

Design and Development of a Desktop-Based Cashier Application at the Legowo Satay Restaurant

Nila Nur'alifah — 2026-06-15

The advancement of information technology encourages various business sectors to adapt, including in managing sales transactions. Resto Sate Legowo, as a local culinary business, still relies on manual recording for cashier operations, which is prone to errors and takes considerable time. To address this issue, this study aims to design and develop a desktop-based cashier application to help accelerate and simplify the transaction process at Resto Sate Legowo. The development method used in this system is Rapid Application Development (RAD), with Unified Modeling Language (UML) as the design tool. This method allows for faster development by actively involving users in each stage of the process. The application is built using the Java programming language and MySQL as the database to store transaction and menu data. The result of this study is a cashier application capable of recording sales transactions, printing receipts, and generating daily reports automatically. System testing shows that the application functions properly and provides convenience for staff in managing transactions. With this application, it is expected that operational efficiency at Resto Sate Legowo can be significantly improved.

Implementation of the RAD Method in the Design and Development of the Melati Library Information System Tebat Agung Village Web-Based

Davena Nola — 2026-06-15

The development of information technology offers solutions for library management, including at Melati Library in Tebat Agung Village, Muara Enim Regency, South Sumatra, which still uses a manual system. Issues include recording, book borrowing and returning, searching, and report generation, all of which are inefficient. This study aims to design a web-based library information system using the Rapid Application Development (RAD) method, which enables fast and structured system development. The system was developed using a descriptive qualitative approach, UML modeling, PHP programming language, and MySQL database. The resulting system simplifies book data management and is expected to improve the efficiency and quality of library services overall.

Web-Based Design and Construction of HMI Member Registration for Prabumulih City

Krisna — 2026-06-15

The Islamic Students Association (HMI) Prabumulih Branch still carries out its member registration process conventionally, such as using physical forms and manual data collection, which can lead to delays and disorganized data management. This study aims to design and develop a web-based member registration system that simplifies administrative processes and improves committee work efficiency. The system development methodology used in this study is the Waterfall method, which includes requirement analysis, system design, implementation, testing, and maintenance. The result of this research is a web-based registration system that can be accessed online by prospective members and the organizing committee, featuring digital form filling, document uploads, and member data management. This system is expected to support a more structured, efficient, and modern registration process within the HMI Prabumulih Branch organization.

Design of a Web-Based Digital Archive Information System for Incoming and Outgoing Mail Case Study of the Rambang Kapak Tengah District Office

Sagita Vira — 2026-06-15

Well-organized archives are an important element in supporting smooth operations and public services in government agencies. The Rambang Kapak Tengah Subdistrict Office still uses manual methods in managing records. This is often inefficient and susceptible to damage and data loss. This research aims to design a web-based digital archive information system that is able to increase efficiency and reliability in managing archives in the office. The method used is the RAD development model which includes requirements analysis, design, implementation and testing stages. The resulting system makes it easy to store, search and manage archives digitally. This helps minimize the risk of data loss and improves ease of access and security of information. It is hoped that the implementation of this system can support more effective and efficient archive management at the Tanjung Rambang village sub-district office.

Development of a Website-Based Cashier Application at the NBO Prabumulih Store Using the RAD Method

Lovita Reira Rambayu — 2026-06-15

NBO Prabumulih Store is a store that sells various kinds of men's and women's clothing, NBO Prabumulih Store is located in North Prabumulih District, precisely in Anak Petai Village. Currently, NBO Prabumulih Store does not have a cashier application for managing payment transactions, so the purpose of this study is to build a cashier application to facilitate payment transactions at NBO Prabumulih Store. This research method uses a qualitative descriptive method with data collection techniques in the form of observation, interviews and literature studies, data sources consist of primary data and secondary data, while for the system development method using the RAD (Rapid Application Development) method, the system design tool used is UML (Unified Modeling Language). The design of this application uses the PHP (Processor Hypertext) programming language, MySQL Database and Coding using Visual Studio Code

Analysis of Hate Speech Againts Gojek Drivers using the Naïve Bayes Algorithm on the Facebook Platform

Nazwa Putri Ananda — 2026-06-15

Social media has become a space where many individuals express their opinions freely, including negative comments that may lead to hate speech. One group often targeted by such speech is Gojek drivers. This study aims to classify user comments on Facebook into two sentiment categories: positive and negative, with a primary focus on negative comments. The data was collected from public Facebook posts using the APIFY scraping tool. After the data was gathered, several preprocessing stages were carried out, including case folding, cleaning, tokenization, normalization, stopword removal, and stemming. The text data was then converted into numerical form using CountVectorizer. The classification algorithm used in this research is Naive Bayes with the MultinomialNB model, as the input data consists of word frequency. The results of the model evaluation show that this algorithm performs well in classifying negative comments, especially in identifying word patterns that commonly appear in hate speech directed toward Gojek drivers.

Sentiment Analysis of Pre-Loved Shoe Product Sales Based on X Reviews with a Comparison of Support Vector Machine (SVM) and Long Short-Term Memory (LSTM) Algorithms

Setyo Harry Nugroho — 2026-06-15

The rapid growth of social media enables consumers to express opinions about products openly, including preloved shoes. These reviews are crucial as they can influence purchase intentions and brand perception. This study aims to analyze user reviews on the X (Twitter) platform using Support Vector Machine (SVM) and Long Short-Term Memory (LSTM) algorithms. A total of 1,005 reviews were collected, then preprocessed and balanced into 738 data consisting of positive and negative sentiments. The results show that SVM achieved an accuracy of 68%, while LSTM obtained 61.49% in its best configuration. Thus, SVM demonstrates better efficiency in classifying simple text, whereas LSTM requires more complex parameters to achieve optimal performance. This research is expected to serve as a reference for utilizing sentiment analysis to support business decision-making in the preloved product market.

Web-Based Goods Inventory Information System Using the Rapid Application Development Method (Case Study: SMK Fatahillah Cileungsi Bogor)

Zainal Musthofa — 2026-06-15

Inventory management at SMK Fatahillah Cileungsi Bogor is not yet supported by an integrated information system, so the process of recording incoming and outgoing goods has not been running optimally. This condition causes difficulties in data management, increases the risk of inventory information discrepancies, and limitations in monitoring real-time availability of goods. This study aims to design and implement a web-based inventory information system that is able to automate and centralize school asset management. The system development method used is Rapid Application Development (RAD), which includes the stages of requirements planning, system design, prototype construction, testing, and implementation. The developed system provides features for managing master data on goods, user management, recording incoming and outgoing goods transactions, borrowing and returning goods, and automatic generation of inventory reports. Test results show that the system can function well according to user needs. The implementation of this web-based information system has been proven to be able to improve data accuracy, accelerate the information search process, and optimize the effectiveness of asset management at SMK Fatahillah Cileungsi Bogor.

Web-Based Cooperative Management Information System Using Agile Method (Case Study: Bungah Bareng Mandiri Banyumas)

Nurul Khikam — 2026-06-15

The development of information technology encourages cooperatives to manage operational activities in an integrated manner to improve the effectiveness and accuracy of data management. The Bungah Bareng Mandiri Banyumas Cooperative still faces obstacles in managing cash and credit transactions, recording customer data, installment payments, and preparing reports because the system used is not yet fully integrated. This study aims to design and implement a website-based cooperative management information system that is able to integrate all operational data into a centralized platform. The research method used is applied research with a qualitative-descriptive approach, while system development is carried out using the Agile method to accommodate user needs iteratively. The implementation results confirm that the improved system is able to increase the regularity and accuracy of data processing, facilitate monitoring of installment due dates, and support the preparation of operational reports more quickly and structured. Thus, the implementation of this information system contributes to increasing the effectiveness, efficiency, and transparency of the operational management of the Bungah Bareng Mandiri Banyumas Cooperative.

Classification of Herbal Leaves Using Support Vector Machine (SVM)

Yakub Takandiwa Takandiwa — 2026-06-15

Indonesia is a country with high biodiversity, including various types of herbal leaves with potential use as traditional medicine. Manual identification of herbal leaves often encounters challenges due to morphological similarities among species and the limited availability of experts, thereby necessitating a fast and accurate technology-based classification method. This study aims to classify 10 types of herbal leaves using the Support Vector Machine (SVM) algorithm with a Radial Basis Function (RBF) kernel. The dataset consists of 3,500 leaf images (350 images per class), from which color features (HSV), texture features (Gray Level Co-occurrence Matrix/GLCM), and shape features (area, perimeter, and aspect ratio) were extracted. The research process includes preprocessing, feature extraction, data splitting into training and testing sets, model training, and performance evaluation. Evaluation was conducted using a confusion matrix, with accuracy as the primary metric due to the balanced class distribution. Precision, recall, and F1-score were employed as supporting evaluation metrics. The results indicate that the SVM model with an RBF kernel successfully classified the 10 types of herbal leaves with an accuracy of 81.29%. Based on per-class analysis, the highest performance was achieved in the Papaya class with an F1-score of 90.00%, followed by Jambu Biji (89.36%) and Pandan (87.14%). In contrast, the lowest performance was observed in the Aloe Vera class with an F1-score of 65.71% and Lime with 70.00%. The model achieved an average precision of 81.16%, recall of 80.73%, and F1-score of 80.94%. Misclassifications primarily occurred among classes with high morphological similarity, such as Aloe Vera, which was frequently misclassified as Pandan (9 cases) and Basil (5 cases). The system has been implemented as a Graphical User Interface (GUI) application that allows users to upload leaf images and obtain classification results along with information regarding their herbal benefits within 1–2 seconds.

Implementation of Agile Method in Village Web Development (Case Study: Tapus Village)

Rahma Pita Kurnia — 2026-06-15

The advancement of information technology has had a significant impact on various aspects of life, including governance at the village level. However, Tapus Village still relies on traditional methods such as bulletin boards, community meetings, and WhatsApp groups to disseminate information to its residents. These methods have limitations in terms of reach, effectiveness, and information archiving. To address these issues, this study aims to develop a village website for Tapus Village using the Agile Software Development methodology. Agile was chosen due to its flexible and adaptive approach, allowing the system to evolve in response to users’ changing needs. The outcome of this research is a village website capable of presenting activity updates and important announcements in a structured and easily accessible format. With the implementation of this website, Tapus Village is expected to enhance transparency, improve service efficiency, and increase community engagement in village development. Moreover, this study demonstrates that the Agile method is effective for developing web-based information systems in village government environments with limited resources.

Implementation of the RAD Method in the Development of a Web-Based Jiwa Baru Village Profile Information System

Firda Fatrika — 2026-06-15

The use of information technology in various aspects of life including government is increasingly important, one of which can feel the positive impact of the application of technology is the village government. In Jiwa Baru Village, the delivery of information is still delivered through speakers or bulletin boards. The limitations in delivering this information often make it difficult for people to get accurate and up-to-date information. The purpose of this study is to design a web-based Jiwa Baru Village Profile Information System which aims to deliver information to the public effectively and efficiently. The method used in this study uses the RAD (Rapid Application Development) method. The results of the evaluation show that all functions of this system run well and in accordance with user needs, with this system it can make it easier for people to find information about the village.

Development of an Ordering Application for Ayek Jerangan Shop in Tanjung Bunut Village Using the Waterfall Method

Dili Muhamad Padholi — 2026-06-15

The development of information technology has improved efficiency in various sectors, including clean water distribution. In Tanjung Bunut Village, ayek jerangan (boiled or refillable drinking water) is a primary water source, but its ordering and distribution system remains conventional, causing irregular supply and fake orders. This study aims to design a web-based ordering application to improve efficiency, transparency, and accessibility for residents. The application is expected to help manage demand and distribution in a more structured manner and enhance service quality for the community.

Website-Based Design of Academic Information System at Rama Chindo PAUD

Mangie Syakila Putri — 2026-06-15

In line with the rapid development of technology and information, the demand for quick, accurate, and efficient information systems is growing, especially in the field of education. PAUD Rama Cindo in Prabumulih currently relies on conventional methods for managing academic data, such as grades, attendance, and student progress reports, which are recorded manually in large books and stored in physical files. This situation leads to various challenges, such as difficulties in accessing data, the potential for document loss, and inefficient information dissemination. This study aims to design and develop a website-based academic information system for PAUD Rama Cindo to enhance efficiency in managing academic data and delivering information to teachers, students, and parents. The developed system will enable digital data management, making academic information easily accessible, quick, and accurate. With this system, it is expected that academic data management can be done in an integrated and more efficient manner, while also introducing PAUD Rama Cindo to a wider community through digital media. The research methods include data collection, system design, and testing of the developed academic information system application. This study provides practical contributions to PAUD Rama Cindo in improving the quality of academic services through the application of information technology, while also offering an academic reference for the development of similar information systems in early childhood education.

Web-Based Futsal Field Reservation Application at Futsal NR Prabumulih

Farras Zain — 2026-06-15

The development of digital technology encourages the sports sector to improve services. Futsal NR Prabumulih still uses a manual booking system that is less practical, especially in terms of time and access. This study aims to design a web-based futsal field booking application to make it easier for users to view schedules, book futsal fields, and make payments online. With a user-friendly interface, this application allows customers to make reservations at any time. In designing this system, UML (Unified Modeling Language) modeling tools such as use case diagrams, activity diagrams, and class diagrams are used to explain system functions, activity flows, and data structures. The system development method used is Rapid Application Development (RAD), which emphasizes the speed of development through iterative prototyping and active user involvement in every stage of development. Testing of the system shows an increase in efficiency in booking management and an increase in customer satisfaction. This application is expected to support modern and integrated services at Futsal NR Prabumulih.

Implementation of the Rapid Application Development Approach for the Academic Information System at Muzakkir Islamic Primary School Prabumulih

Ardi Ardiansyah — 2026-06-15

This study stems from the limited application of information technology, which continues to influence various aspects of human life—particularly in the fields of employment, business, and, more specifically, at Muzakkir Islamic Elementary School in Prabumulih, where data management is still handled through conventional methods. Data for this research was collected using a descriptive method with a qualitative approach, employing observation, interviews, and literature review. The design of the academic information system applied the Rapid Application Development (RAD) methodology to assist developers in analyzing and designing the system based on actual needs. The Unified Modeling Language (UML) served as a tool for designing the school’s academic information system, supporting the developers during the process. The developed system was implemented using the Hypertext Preprocessor (PHP) programming language and a MySQL database.

Development of Web System for Sales Optimization at CV. CS Swalayan using Association Rule Method

Steven Imanuel Naibaho — 2026-06-15

CV. CS Swalayan encounters challenges related to declining consumer purchasing power and the underutilization of transactional data for analyzing customer purchasing patterns. This study aims to develop a web-based system employing Association Rule methodology with the Apriori algorithm to optimize sales performance, identify top-selling products, and determine frequently co-purchased product combinations. The research methodology encompasses the collection of 296 sales transaction records for basic commodity products from CV—CS Swalayan during January 2025, followed by data preprocessing procedures. The Apriori algorithm is implemented with minimum support and confidence thresholds set at 0.01 and 0.3, respectively. The web-based system is developed using Python with the Flask framework for backend functionality, MySQL for database management, and validated through black-box testing methodology. The findings reveal the generation of 14 valid and robust association rules, notably "if Selai Srikaya Ngetop is purchased, then Roti Tawar Kupas Ngetop will be purchased" (confidence: 100%; lift ratio: 49.3) and "if Beras Sukaraya Cap Gurih 10KG is purchased, then Minyak Kita Minyak Goreng Sawit 1ltr will be purchased" (confidence: 100%; lift ratio: 16.4). The developed web system successfully passed black-box testing with a 100% success rate. This research contributes by providing a system that enables CV. CS Swalayan will make data-driven decisions to optimize sales strategies, marketing approaches, and inventory management practices.

Web-Based Operational Management Information System for Prospective Indonesian Migrant Employees Using Agile Method (Case Study: PT. Bahana Mega Prestasi Bekasi)

Aldi Jaya Mulyana — 2026-06-15

Bahana Mega Prestasi, an Indonesian Migrant Worker Placement Company (IMWPC), faces challenges in managing operational data for Prospective Indonesian Migrant Workers (PIMW). This is due to the lack of integration of registration, attendance, and eligibility assessment processes within a single information system. This situation increases the administrative burden and the potential for data inconsistencies. This research aims to design and implement a web-based PIMW Operational Management Information System capable of centrally integrating all administrative processes. The system was developed using Agile methods with an iterative approach to ensure the system meets user needs. The system was built using the PHP programming language with the Laravel framework and a MySQL database. Implementation results indicate that the developed system is able to support the PIMW registration process, attendance monitoring, and candidate eligibility evaluation in a more structured and real-time manner. The implementation of this system contributes to improving the orderliness of data management, supporting managerial decision-making, and increasing operational efficiency at PT. Bahana Mega Prestasi.

Comparison of Naive Bayes and KNN Algorithms for Heart Attack Disease Classification

Syahril Arsad — 2026-06-15

This Heart attack is one of the leading causes of death worldwide and requires early diagnosis to reduce fatal risks. This study aims to compare the performance of the Naive Bayes and K-Nearest Neighbors (KNN) algorithms in classifying heart attack disease. The dataset used consists of medical records containing clinical parameters such as age, blood pressure, cholesterol level, and heart rate. The research methodology includes data preprocessing, splitting the dataset into training and testing sets, and evaluating performance using accuracy, precision, recall, and F1-score metrics. The results show that Naive Bayes demonstrates advantages in computational speed and performs well on smaller datasets, achieving an accuracy of 85%. In contrast, KNN provides better performance on larger datasets, reaching an accuracy of 90%, particularly when the optimal K value is applied. These findings indicate that algorithm selection for heart attack classification depends on dataset characteristics and specific implementation needs. This study is expected to contribute to the development of artificial intelligence–based clinical decision support systems for early heart attack diagnosis and improved healthcare outcomes.

Implementation of Deep Learning Based on Convolutional Neural Network for Detecting Images of Solar Panel Damage in Smart Grid Systems

Camelia Putri Lestari — 2026-06-15

This study aims to implement Deep Learning based on Convolutional Neural Network (CNN) in detecting solar panel damage using thermal images as part of a Smart Grid system. The main problem addressed is the difficulty of early automatic identification of solar panel cell damage using conventional methods. Through the CNN approach, this study developed a classification model to distinguish between damaged (Defective) and undamaged (Non-Defective) solar panel conditions. The research stages included thermal image dataset collection, pre-processing, model training, and performance evaluation. The results showed that the CNN model was able to achieve an accuracy of over 87% with stable performance on the validation data. Visualization using the Grad-CAM method helps interpret the damaged areas that are the focus of the model's decision.

Sentiment Analysis of Social Media X Users Toward Finance Minister Purbaya Yudhi Sadewa Using the Support Vector Machine Algorithm

Adian Fahreza Surbakti — 2026-06-15

In this digital era, the rapid advancement of information and communication technology has transformed social media platforms particularly X (formerly Twitter) into a primary space for public discourse concerning government policies. The Minister of Finance, Purbaya Yudhi Sadewa, has become a focal point of public debate, garnering reactions ranging from appreciation to criticism regarding his management of national finances. However, manual sentiment analysis is impractical, time-consuming, and prone to subjectivity when handling the massive and continuously expanding volume of social media data. Therefore, an automated, machine learning-based approach is essential to process this big data into strategic insights for mapping public sentiment. This study aims to objectively analyze public sentiment toward the Minister of Finance by implementing the Support Vector Machine (SVM) algorithm within the CRISP-DM (Cross-Industry Standard Process for Data Mining) framework. The methodology includes data crawling, text preprocessing, and feature extraction using the TF-IDF (Term Frequency – Inverse Document Frequency) method. Analysis of 3,927 tweets reveals that public opinion is dominated by negative sentiment at 54.2%, followed by positive sentiment at 36.9% and neutral sentiment at 8.9%. The developed SVM model achieved a classification accuracy of 72.43%, demonstrating that this machine learning approach is both effective and reliable for mapping public perception. These findings indicate that the Minister of Finance, Purbaya Yudhi Sadewa, faces significant public scrutiny, and this data-driven analysis serves as a strategic tool for evaluating the policies under his administration.

Failure Analysis of Switching Scheme Failures in Loop Protect Multiplexer Telecommunication Networks at PT. PLN (Persero) UP2B DKI Jakarta & Banten

Rizki Dwi Dermawan — 2026-06-15

PLN (Persero), through UIP2B JAMALI, relies on a loop-topology Loop Protect Multiplexer as its telecommunications backbone to support real-time SCADA, VoIP, and protection services. However, from 2022 to 2024, 36 switching failure incidents occurred in UP2B DKI Jakarta–Banten. This study analyzes the root causes, operational impacts, and recommendations for continuous system reliability improvement. The research employs a case study method and the PPDIOO approach to examine switching failures of the Loop Protect Multiplexer in UP2B DKI Jakarta–Banten. Data were collected through observations of fault history (2022–2024) and interviews. BER testing and QoS parameters refer to the ITU-T Y.1564 standard to formulate recommendations for improving the reliability of the 150 kV backbone network. Testing results indicate that under normal conditions, the system meets SLA requirements in accordance with ITU-T Y.1564, with stable throughput and zero frame loss. However, when one link fails, frame loss occurs during switching despite stable throughput, resulting in SLA failure. The root cause lies in a reactive and non-seamless switching mechanism, creating cross-layer impacts on critical services within PT. PLN (Persero).

Application of the K-Means Clustering Algorithm in the Analysis of Popularity and Growth Trends of Python Packages on the PyPI Dataset

Muhammad Rafli Wijaya — 2026-06-15

The rapid growth of the Python ecosystem has led to an increasing number of packages on the Python Package Index (PyPI), generating a massive volume of download data. This data can be utilized to analyze popularity levels and growth trends of libraries used by the developer community. This study aims to identify popularity patterns and growth trends of Python packages using the K-Means Clustering algorithm. The dataset was obtained from PyPI via the Google BigQuery platform with a one-year observation period using a 1% sampling technique. The pre-processing stage included a filtering process to select the 100 packages with the highest number of downloads and the formation of six main features representing the characteristics of library usage patterns. The data was then normalized using Standard Scaling, while the optimal number of clusters was determined using the Elbow Method and evaluated using the Davies-Bouldin Index (DBI) and Silhouette Score. The results showed that the optimal number of clusters is four, with a DBI value of 0.5534 and a Silhouette Score of 0.5748 (the highest among k = 2-10 ), representing the categories of ecosystem foundation libraries, medium-popularity libraries, libraries with concentrated download spikes, and libraries with very rapid usage growth. These results indicate that K-Means Clustering is effective for identifying popularity patterns and library growth trends in large-scale PyPI datasets.

Application of K-Means Clustering: Bot Activity and Sybill Attack Detection on the Solana Blockchain

Bryant Tinambunan — 2026-06-15

With the development of Blockchain technology, for example, the Solana Blockchain has generated enormous amounts of data and possesses the 5Vs of Big Data: volume, velocity, value, veracity, and variety. This has brought challenges, for example, in distinguishing transactions carried out by humans from automated bots that often carry out market manipulation or Sybil attacks. Therefore, this research aims to detect bot activity on the Solana network by applying data mining techniques, namely the K-Means Clustering algorithm. From the large transaction data that will be extracted only a portion from the public Solana dataset in BigQuery, it will then be processed through a preprocessing stage to normalize the data and simplify complex data into simpler variables before being grouped. Because the extracted data is in the form of unlabeled data groups (unsupervised data), the Clustering Method is used because of its ability to recognize data groups based on behavioral or characteristic similarities without requiring initial data labels (unsupervised learning). The main variables used for the grouping process include transaction frequency, inter-arrival time (inter-transaction), and the number of unique program interactions. The results of this analysis are expected to map transaction accounts into several clusters based on their transaction patterns, allowing for the classification of bots and humans. This research is expected to demonstrate that Big Data infrastructure such as Google Cloud, using data mining techniques (Clustering), can be used to maintain the security and integrity of the blockchain ecosystem.

Implementation of the Heuristic Evaluation Method in the Design of the School Academic Information System Website

Michelle Francisca — 2026-06-15

In the world of education, the role of teachers and parents is very influential in the process of improving student learning achievement. However, in reality, most parents only give responsibility to teachers at school to improve student learning achievement. Parents of students rarely monitor the development of their children's learning abilities due to the lack of information about it. Global Prima National Plus School is one of the leading private schools in Medan, located on Jalan Brigjend Katamso. Currently, Global Prima National Plus School uses Microsoft Excel to manage student data and student test scores. However, the implementation of this system still has several weaknesses, namely parents cannot monitor attendance and directly know the development of student scores and behavior. This will reduce parental participation in their children's educational development. To solve the problems faced by Global Prima National Plus School, an application can be created to monitor student learning development. By using this application, parents of students can obtain information about student attendance data, attitude and behavior scores, assignment scores, and test scores directly, without having to wait for report cards to be distributed. With this web-based student learning progress monitoring application, parents can find out information about student attendance, attitude scores, behavior, exam scores and assignment scores which can be accessed directly through the school website.

Application of K-Means Clustering for Urban Transportation Pattern Analysis Using Big Data Trip Dataset

Tegas Ramadhan — 2026-06-15

The rapid growth of urban transportation systems has led to the generation of massive amounts of data, commonly referred to as big data. This study aims to analyze transportation patterns using large-scale data obtained from the NYC Taxi Trip Records. The dataset exhibits key big data characteristics, including volume, velocity, and variety. This research applies the K-Means clustering algorithm to group taxi trip data based on features such as trip distance, fare amount, and trip duration. Several preprocessing techniques are performed, including data cleaning, feature engineering, sampling, and normalization. The optimal number of clusters is determined using the Elbow Method and Silhouette Score. The results show that the dataset can be effectively grouped into three clusters representing distinct transportation patterns. These findings demonstrate the capability of clustering techniques in extracting meaningful insights from large-scale datasets and highlight their potential application in urban transportation planning.

Development of a Web-Based System for Recording and Reporting Palm Weights Using Laravel at PT. Graha Prima Lestari

Fredynand Marcos — 2026-06-15

This research was initiated by operational problems in the palm oil weighing process, which was conducted manually. The manual method often caused calculation errors, delays in making report , and risk of data loss. To address these issues, a web-based palm oil weighing application was developed using the Laravel framework and the Waterfall development method, supported by a relational database to manage data in an integrated manner.

The application implements a role-based access system to manage permissions for administrators, weighing operators, and management. The system records gross weight, tare weight, and automatically calculates net weight while generating accurate reports efficiently. With this system, the weighing process is expected to become more efficient , precise and structured.

Designing a Web-Based Financial Information System at GKS Palindi using the Rapid Application Development Method

Serlince Pindi Kualak — 2026-06-15

In the rapidly evolving information age, technology plays an important role in improving the efficiency of data management, including in church institutions. The Church, as a religious institution with an important role in the spiritual and social life of the people, often faces challenges in financial management, especially in recording congregation donations, operational expenses, and making transparent and accurate financial statements. Good financial management is needed to ensure accountability and transparency, as well as facilitate reporting to the congregation and other related parties. GKS Palindi, a church in Kawangu District with 335 congregations, faces problems in the financial recording system that is still manual using books. This causes data corruption, which risks disrupting the smooth flow of the financial management process. Therefore, this church needs the implementation of a technology-based information system that can facilitate the recording and management of financial transactions efficiently, especially for the six main posts: tithe, thanksgiving, part (household worship), monthly dependents, offerings, and miscellaneous posts. With this system, the church can reduce the potential for human error, monitor cash flow more easily, and provide more accurate and timely financial reports. The right information system can help GKS Palindi in maintaining the continuity of church operations and increasing the congregation's trust in the transparency of financial management. The system development method used is Rapid Application Development (RAD), which allows the creation of a system quickly and responsively to user needs. The implementation of a technology-based recording system is expected to overcome existing problems, as well as support the smooth running of church activities in the long term.

Clusterization of Family Planning Participants Based on Pregnancy Risk Using K-Means Algorithm in Ciherang Village

Melva Regina Arpratika — 2026-06-15

This study aims to group family planning (KB) participants in Ciherang Village based on pregnancy risk levels using the K-Means clustering algorithm. The identification of pregnancy risk is still performed manually, resulting in less effective analysis. Therefore, a data mining approach is applied to improve decision-making accuracy.

The data used in this study were obtained from KB cadres, including variables such as age, number of children, education, occupation, and contraceptive methods. The research method follows the Knowledge Discovery in Database (KDD) stages: data selection, preprocessing, transformation, data mining, and evaluation. The K-Means algorithm is used for clustering, while the Davies–Bouldin Index (DBI) is applied to evaluate clustering quality.

The results show that the optimal number of clusters is K = 2 with a DBI value of 0.721. The first cluster represents low pregnancy risk participants, while the second cluster represents high pregnancy risk participants. Age and number of children are identified as the most influential factors.

This study provides useful insights for healthcare providers in developing targeted strategies for family planning programs.

Keywords: Data Mining; Davies–Bouldin Index; K-Means Clustering; Pregnancy Risk; Family Planning

Prediction of Peritonitis Infection Risk in CAPD Patients using Random Forest Algorithm

Silviani Gustaman — 2026-06-15

Peritonitis is a serious complication frequently experienced by patients undergoing Continuous Ambulatory Peritoneal Dialysis (CAPD) and may worsen patient outcomes if not detected early. This study aims to develop a machine learning model to predict peritonitis risk using the Random Forest algorithm and to interpret prediction results using Explainable Artificial Intelligence (XAI). The study utilized a secondary dataset obtained from Kaggle consisting of 20,538 clinical records that were transformed to represent CAPD-related clinical parameters. The research stages included data preprocessing, feature selection using SelectKBest (f_classif), dataset splitting into training and testing sets, model development using Random Forest, and performance evaluation using accuracy, precision, recall, F1-score, and Area Under Curve (AUC). Model interpretability was analyzed using SHAP to identify feature contributions. The experimental results demonstrate that the proposed model achieved an accuracy of 98.70%, precision of 98.22%, recall of 99.24%, F1-score of 98.73%, and AUC of 1.00. The findings indicate that Random Forest provides highly reliable predictive performance and interpretable insights into clinical features influencing peritonitis risk. The developed model has potential to support clinical decision-making systems for early detection of peritonitis risk in CAPD patients.

UI/UX Design of Laundry Pick-Up and Delivery Application using Prototyping Method

Margaretha Natalia Simamora — 2026-06-15

Digital transformation demands operational efficiency in the conventional laundry industry, which is still hampered by manual management and limited geographic reach.In response to this phenomenon, this research focuses on developing a UI/UX design for the Laundry Express mobile application with superior pickup & delivery service features. The main goal is to reduce the potential for data input errors while providing information transparency for users. Through an iterative prototyping method, the design process includes needs identification and continuous evaluation using Figma. The final product, a high-fidelity prototype, integrates order tracking features, automatic cost calculation based on weight, and a service assessment module. Evaluation using Likert scale for usability measurement demonstrated a high level of ease of navigation, allowing users to complete transactions without technical obstacles. This study concludes that the iterative prototyping approach is effective in producing intuitive application designs that meet the needs of modern society who require flexible laundry services.

Analysis of Student Errors in Solving Problems Involving Curved-Surface Geometric Shapes Based on Newman’s Error Analysis

floricytha sihombing — 2026-06-15

This study aims to evaluate the errors made by students when solving problems involving three-dimensional shapes with curved sides, using Newman’s error analysis approach. The research employed a descriptive qualitative method and was conducted at Imelda Private Junior High School in Medan during the second semester of the 2025/2026 academic year, with the participation of 18 students selected through purposive sampling. Data collection tools consisted of written tests, interviews, and documentation. Data analysis was conducted by referring to Newman’s five stages of error: reading, comprehension, transformation, process skills, and coding.

The research findings indicate that the most common errors were process skill errors, accounting for 25.9%, and transformation errors, accounting for 20.3%, while errors in the reading, comprehension, and coding stages were not identified. Students with low ability typically struggle to find the formula and proceed with the problem-solving process; students with moderate ability tend to make errors during calculations; whereas high-ability students successfully solve problems accurately and in an organized manner.

Thus, it can be concluded that most student errors are caused by an inability to select the correct formula and a lack of precision during calculations. Therefore, it is crucial to implement teaching methods that focus on conceptual understanding and procedural skills to minimize the errors students make when tackling mathematical problems.

Decision Support System Using the Analytical Hierarchy Process Method in Determining Credit Recipient Eligibility

Erika Nia Devina Br Purba — 2026-06-15

Banks play a fundamental role in improving public welfare by collecting funds through savings and redistributing them as credit. Although credit is the primary source of bank revenue, it carries significant risks if the feasibility analysis of prospective borrowers is flawed, potentially leading to non-performing loans that disrupt financial stability. BPR Nusantara Bona Pasogit 17 faces this challenge as it currently lacks an automated decision support system, resulting in assessments that are often inconsistent or subjective. This research aims to develop a web-based decision support system using the Analytical Hierarchy Process (AHP) method to determine credit recipient eligibility. Developed using PHP and MySQL, the system incorporates criteria management, AHP calculation processing, and automated eligibility ranking. Comprehensive validation through black-box and white-box testing confirmed that all functional components performed correctly with consistent "PASS" results. The AHP implementation produced a Consistency Ratio (CR) of 0.03797, indicating high reliability in decision-making. Criterion priority weights were identified as: Income (0.386), Character (0.219), Loan Amount (0.162), Collateral (0.103), Loan Term (0.07), and Age (0.06). System testing on 100 customer records resulted in a maximum eligibility score of 0.93501 and a minimum of 0.41839.

Developing a Web-Based E-Commerce Application for Toko Oleh-Oleh Khas Prabumulih

Ivan Mei Dwintara — 2026-06-15

Prabumulih Typical Souvenir Shop is a business unit selling various pineapple-based processed products that still faces constraints in promotional reach and manual transaction efficiency. This study aims to design and build a web-based e-commerce information system as a solution for digital marketing and sales. The system development method used is Rapid Application Development (RAD), consisting of requirements planning, user design, construction, and cutover phases. The application was built using PHP programming language and MySQL database. The results show that the application successfully facilitates online transactions, real-time stock management, and automated sales reporting, which significantly enhances the shop's operational efficiency.

Implementation of Convolutional Neural Network for Emergency Sound Detection for Hearing-Impaired Individuals on Android

Muhammad Akram Fais — 2026-06-15

Hearing impairment is a condition characterized by partial or total loss of hearing ability, which may occur congenitally or be caused by factors such as injury, disease, or prolonged exposure to excessive noise. This study aims to develop an Android-based emergency sound detection system using the Convolutional Neural Network (CNN) method. The research workflow includes problem identification, data collection, data preprocessing, CNN model training, model evaluation, Android application development, and system testing. Experimental results show that the best-performing model achieved an overall accuracy of 93%. The trained model was then implemented into an Android application to enable real-time sound classification and to provide visual notifications when emergency sounds are detected. The evaluation results indicate that the CNN model is capable of accurately classifying emergency sounds and operates effectively on Android devices.

Eye Disease Classification System Based on Fundus Images Using the InceptionV3 Architecture

Annisa Aulia — 2026-06-15

This study aims to develop an automated eye disease classification system based on retinal fundus images using the InceptionV3 deep learning architecture. The dataset consists of four classes: cataract, diabetic retinopathy, glaucoma, and normal, collected from public sources and clinical data. The proposed method applies several preprocessing techniques, including background segmentation, data augmentation, data normalization, and an 80:20 data split to improve model performance and generalization. Transfer learning is implemented by utilizing pretrained ImageNet weights and modifying the final layers to suit the classification task. The model is trained using the Adam optimizer with a learning rate of 0.001 and categorical cross-entropy loss function. Evaluation results show that the model achieves an accuracy of 96%, with average precision, recall, and F1-score values of 0.97, 0.96, and 0.97, respectively. The confusion matrix analysis indicates that most predictions are correctly classified, demonstrating strong performance across all classes. Furthermore, the model is successfully integrated into a web-based system that enables users to upload fundus images and obtain classification results automatically. These findings indicate that the proposed system can effectively assist in early detection of eye diseases and support clinical decision-making.

UI/UX Design of an Incoming and Outgoing Mail Information System using the Design Thinking Method

Nurhayati — 2026-06-15

Kantor Inspektorat Kota Jambi saat ini mengelola surat masuk dan keluar secara manual menggunakan buku catatan, yang menyebabkan risiko kehilangan dokumen dan inefisiensi dalam pengambilan data. Penelitian ini bertujuan untuk mendesain Antarmuka Pengguna (UI) dan Pengalaman Pengguna (UX) untuk Sistem Informasi Surat Masuk dan Keluar (SIMAK) yang ramah pengguna untuk meminimalkan kendala operasional tersebut. Metode yang digunakan adalah Design Thinking , yang terdiri dari lima tahapan: empati, definisi, ide, prototipe, dan pengujian. Hasil penelitian ini adalah prototipe aplikasi berbasis web yang dirancang menggunakan Figma, yang menampilkan alat manajemen untuk surat masuk, surat keluar, disposisi, dan laporan. Pengujian yang dilakukan menggunakan metode System Usability Scale (SUS) dengan 10 responden menghasilkan skor rata-rata 89,75 . Skor ini menempatkan desain aplikasi dalam kategori " Diterima " dengan peringkat " Baik ", yang menunjukkan bahwa sistem mudah dipahami dan secara efektif memenuhi kebutuhan pengguna.

Analysis of Mathematics Percentage Calculation Strategies Quickly and Accurately Based on a Literature Review

Adi Sinaga — 2026-06-15

Percentages are one of the important concepts in mathematics that are widely used in various contexts of daily life such as economics, commerce, and decision-making. However, various studies show that the concept of percentage is still a material that is quite difficult for students and prospective mathematics teachers to understand. These difficulties are generally related to the understanding of the basic concept of percentage, the use of the percent symbol (%), and the ability to relate the percentage value to the reference value in a problem. This study aims to analyze various percentage calculation strategies that can be carried out quickly and accurately based on the results of previous research. The method used in this study is a literature study by reviewing various scientific articles from SINTA-accredited national journals and international journals that are relevant to the topic of percentages in mathematics learning. Data is collected through documentation techniques by examining and analyzing research findings related to the percentage calculation strategy. The results of the study show that the use of mental calculation strategies, understanding the basic concept of percentages, and the use of visual models can help improve students' ability to calculate percentages more effectively and efficiently. Therefore, the right calculation strategy is essential to support the understanding of the concept of percentages in mathematics learning.

Sentiment Analysis of Film Audience for IPAR ADALAH MAUT Using Support Vector Machine

Surya Agung Agan Saputra — 2026-06-15

This study aims to analyze user sentiment on social media X (formerly Twitter) toward the film Ipar Adalah Maut using the Support Vector Machine (SVM) method. The data were collected through a crawling process using the snscrape library, focusing on tweets containing keywords related to the film title. The preprocessing stages included data cleaning, case folding, tokenization, stopword removal, and stemming, while feature extraction was performed using Term Frequency Inverse Document Frequency (TF-IDF). Sentiment was classified into two categories, namely positive and negative, using the SVM algorithm. The results showed that the model achieved 100% accuracy on the training data and 82% accuracy on the testing data, indicating good generalization performance, although there is a potential risk of overfitting due to the gap between training and testing results. These findings demonstrate the effectiveness of SVM in analyzing sentiment related to film discussions on social media and provide a basis for future research by incorporating larger and more balanced datasets.

Flood Prediction for the Wampu River Basin Using the Simple Additive Weighting Method:A Case Study of the Wampu River in Bahorok

Miftahul Janna — 2026-06-15

Flood is one of the natural disasters that frequently occurs in the Wampu Watershed (DAS Wampu), especially in Bahorok District. Flood risk is influenced by several factors such as rainfall, slope gradient, land use changes, and river depth. The problem in this study is the absence of a decision support system that can objectively determine flood risk levels. This study aims to determine the criteria and weights of flood risk, apply the Simple Additive Weighting (SAW) method, and analyze the accuracy level of the SAW method in determining flood risk. The method used in this research is the Simple Additive Weighting (SAW) method through several stages including criteria weighting, decision matrix construction, data normalization, preference value calculation, and alternative ranking. The research data consists of 18 villages with four criteria: rainfall, slope gradient, land use change, and river depth. The results show the classification of flood risk levels into high, medium, and low categories based on the obtained preference values. Villages with the highest preference values indicate a higher level of flood vulnerability compared to other villages. The model evaluation results indicate that the SAW method has an accuracy level of approximately 90% in determining flood risk classification. Based on these results, it can be concluded that the SAW method can be used as a decision support system to determine flood risk levels and provide recommendations for priority flood mitigation areas in Bahorok District.

UI/UX Design of an Android-Based Sales Application at Naureen Shop using the User-Centered Design Method

Yessi Hartiwi — 2026-06-15

The development of digital technology has shifted consumer behavior to prioritize convenience and efficiency through online shopping. Naureen Shop, a women's clothing business, currently faces operational constraints due to manual sales, promotion, and data recording processes, resulting in sub-optimal service. This study aims to design the User Interface (UI) and User Experience (UX) of an Android-based sales application for Naureen Shop to enhance business effectiveness. The method employed is User Centered Design (UCD), a design approach focusing on user needs and characteristics through stages of identifying the context of use, specifying user requirements, creating design solutions using Figma, and evaluation. The design testing results using the System Usability Scale (SUS) method with 15 respondents yielded an average score of 77.0. This score indicates that the application design falls into the "Acceptable" category with a "Good" Adjective Rating. Consequently, the resulting design solution fulfills functional aspects and provides a satisfying user experience to support transaction processes at Naureen Shop

Application of the Tsukamoto Fuzzy Inference System Method for Rainfall Prediction in the Adolina Area

Aditia Sanjaya — 2026-06-15

Rainfall is one of the key elements in the climate system that significantly affects various sectors, such as agriculture, spatial planning, and disaster mitigation. Adolina, a region with tropical weather characteristics and highly fluctuating rainfall, requires an accurate prediction system to support informed decision-making. This study applies the Fuzzy Inference System (FIS) Tsukamoto method to predict rainfall based on input variables such as air temperature, humidity, and wind speed. The Tsukamoto method is chosen for its capability to handle uncertainty and produce crisp output values through inference and defuzzification processes based on a set of fuzzy rules. The results show that the Tsukamoto FIS provides reasonably accurate and consistent rainfall predictions with a low error rate. Therefore, this approach can serve as an effective alternative in weather decision-support systems for the Adolina area.

Implementation of Simple Queue and Content Filtering for Bandwidth Management on WLAN and LAN Networks

Zura Permata — 2026-06-15

SMK Negeri 1 Sungai Raya is a vocational high school located on Sultan Agung Street, Kuala Dua, Sungai Raya District, Kubu Raya Regency, West Kalimantan, which offers programs such as Visual Communication Design (DKV) and Broadcasting that utilize the internet to support learning activities including completing assignments, accessing educational resources, and submitting schoolwork; however, problems frequently occur in the laboratory network such as buffering, network downtime, and bandwidth congestion due to simultaneous usage, therefore bandwidth management using the simple queue method was implemented along with content filtering to block access to social media and online gaming websites in order to prevent disruptions to the learning process, and the results showed improvements in network performance where on the LAN network throughput decreased by 0.5735%, packet loss decreased by 0.0969%, delay decreased by 0.5942%, and jitter decreased by 0.9182%, indicating better stability and efficiency, while the WLAN network in the laboratory was also successfully installed, providing improved connectivity and supporting a more effective and focused learning environment.

Design and Construction of a Village Tourism Monitoring and Evaluation System Web Based

Alfian Maulana — 2026-06-15

Development of village tourism in the Special Region of Yogyakarta requires structured and sustainable management, particularly in monitoring and evaluation as a basis for stakeholder decision-making. Current challenges include unintegrated tourism village data, manual evaluation processes, and limited public access to tourism information and services. This study aims to design and develop a web-based monitoring and evaluation system that integrates registration, data management, monitoring, scoring, and information presentation in a centralized platform. The system is developed using the Extreme Programming method, which includes planning, design, coding, and testing stages, with functional testing conducted through Black Box Testing. The technologies used include React JS for the interface, Express JS for the backend, Supabase as the database, and Google Cloud Storage for data storage. The results indicate that all main system features function according to requirements, supporting more effective monitoring and evaluation processes, improving data accuracy, and enhancing accessibility of information for both the public and local government. Furthermore, this system has the potential to serve as a foundation for regional tourism data integration and to support sustainable tourism village development policies, while also contributing practically to improving integrated digital public information services at the national level.

Chinese Script Handwriting Pattern Introduction Application Design with Algorithm CNN-SVM

Jacqueline Kwanori — 2026-06-15

The Chinese script has a high level of visual complexity because each character consists of thousands of intricate strokes. This is a big challenge for second-language learners, especially in recognizing the various variations of human handwriting. This study aims to design an accurate and efficient application for the recognition of Chinese handwriting patterns based on Android using a hybrid model of Convolutional Neural Network (CNN) and Support Vector Machine (SVM). In this system, the CNN works like a human eye that distinguishes the details of the shape of an image, while the SVM serves as the brain that decides what characters are being written. The data used in the training process included 7,330 Chinese characters pulled from the Kaggle platform. The results of the study show that the application was successfully designed and able to display character shapes, how to read (pinyin), and the meaning of words offline without the need for an internet connection. Based on testing the Black Box method, all of the app's features are proven to work validly. The study concluded that the use of the CNN-SVM hybrid model was highly effective in recognizing diverse handwriting variations, although the degree of accuracy remained dependent on the clarity of the quality of the images taken by the user.

Classification of Herbal Plants Based on Leaf Images Using Gray Level Co-Occurrence Matrix and K-Nearest Neighbor

Fahmi Nur Alimsyah Purba — 2026-06-15

Herbal plants have long been used as traditional medicine. However, many people struggle to tell different herbal leaves apart because they look quite similar. This study tries to build a system that can recognize two types of herbal leaves, Moringa and Katuk, simply from their photos. We used GLCM to extract texture features from the leaves, then classified them using KNN. The dataset came from Kaggle, with 480 leaf images in total. Before processing, we cropped the images, resized them to 256x256 pixels, and converted them to grayscale. GLCM features were taken from four angles (0°, 45°, 90°, 135°) and then averaged. This gave us four texture values: contrast, correlation, energy, and homogeneity. We tested KNN with k values from 1 to 15 and five different distance metrics. The best result we got was 94% accuracy, using Manhattan distance with k=1. This system could help everyday people identify medicinal plants more easily without needing lab tests.

Student Mental Health Monitoring System Based on Daily Activities with the SVM Method

Stella Crystal — 2026-06-15

Student mental health is a crucial issue that requires effective and responsive self-monitoring systems. This study aims to develop "LacakJiwa," an Android-based mobile application designed to monitor student mental health through the analysis of daily activity patterns. The method employed is the Support Vector Machine (SVM) with a Radial Basis Function (RBF) kernel to classify mental health risks into low and high categories. Input data includes sleep duration, daily step count, gadget usage, and social interaction duration collected from 146 student data entries. The SVM model is integrated into the application using TensorFlow Lite to enable on-device classification, ensuring user privacy through SQLite local database storage. Testing results on 44 test samples showed an accuracy rate of 52.27%, precision of 36.36%, and recall of 22.22%. While the system was successfully implemented technically, the low recall value indicates significant challenges in detecting complex non-linear behavioral patterns in students. This research provides a foundation for developing digital self-control instruments that are adaptive to Indonesian local culture.

Comparative Analysis of Sobel, Prewitt, and Canny Methods in Detecting Object Edges in Betta Fish Images

Alfin Alfarizi — 2026-06-15

Edge detection is a crucial stage in digital image processing for recognizing the shape and structure of an object. The application of edge detection to betta fish images presents a unique challenge due to their layered, intricately textured, and often semi-transparent fin morphology. This study aims to analyze and compare the performance of three edge detection algorithms, namely Sobel, Prewitt, and Canny, in extracting shape features from betta fish images. The research methodology involved converting the dataset images into a grayscale format and subsequently implementing the three algorithms using the OpenCV library in the Python programming language. The evaluation was conducted visually by observing the sharpness of the edge lines, object continuity, and the occurrence of noise. The results indicate that the Canny algorithm provides the most optimal performance, as it is capable of detecting the thin edge lines of the fish fins with greater detail and continuity due to its hysteresis thresholding process. Meanwhile, the Sobel and Prewitt methods produced thicker edge lines but were less sensitive to the details of the transparent fins. This study is expected to serve as a reference in selecting the appropriate segmentation method for biological objects with complex morphologies.

Smart Absen Implementation of a Facial Recognition-Based Student Attendance System Using the Haar Cascade Method and LBPH

Frengki Alfredo Matondang — 2026-06-15

Manual attendance systems in higher education institutions are often hampered by inefficiency, data inaccuracy, and vulnerability to fraud such as proxy attendance. This study presents the design and implementation of Absen Smart, a face recognition-based attendance system developed using the Haar Cascade and Local Binary Pattern Histogram (LBPH) algorithms within the React.js and Flask frameworks. This system enables the automatic and real-time identification of students via a webcam without requiring additional hardware. Face detection is performed using the Haar Cascade classifier from OpenCV, while face recognition uses the LBPH Face Recognizer with a confidence threshold of 50. Testing was conducted with 28 registered students from the Computer Science Program at UNIMED, Class A, 2024 cohort. Functional evaluation results show that all seven core system features—including face detection, face recognition, duplicate prevention, automatic absence tracking, and Excel report generation—were successfully executed with a 100% success rate. The system achieved a facial recognition accuracy of 92.86%, with an average processing time of 1.2 seconds per verification. These results indicate that the proposed system is an effective, practical, and scalable solution for automating academic attendance in a university setting.

Implementing the Procedural Generation Method for Placing Dynamic Objects in a Roblox-Based Adventure Game

Muhammad Hiszat — 2026-06-15

Procedural generation is a content-creation technique that has become increasingly important in modern game development. However, on the Roblox platform, dynamic object placement still faces challenges such as overlapping, illogical positioning, and blocked navigation paths when relying solely on pure random methods. This research implements the Rule-Based Random Generation algorithm to manage the automatic placement of dynamic objects (enemies, treasure chests, and traps) in a Roblox-based adventure game. The proposed method combines randomization with constraint validation, including boundary check, overlap check using Euclidean distance, restricted zone check, and cross-type relational constraints. The system was developed in Roblox Studio with the Luau scripting language using a prototyping methodology and a modular architecture comprising ObjectSpawner, ConstraintValidator, SpatialGrid, and DungeonGenerator. Functional testing was conducted across 10 game sessions on a 1000 × 1000 studs map with a configuration of 340 enemies, 10 chests, and 50 traps. The results show that the system successfully placed all objects without any constraint violation, produced significant spatial variation between sessions (ranging from 86.31 to 2358.00 studs), and maintained level playability in every session. The average spawning execution time was 336.62 ms per session (0.84 ms per object), demonstrating the computational efficiency of the proposed method.

Recommendation System for Selecting Maternity Hospitals in Pontianak using Weighted Product Method

Ervayana Sari — 2026-06-15

In this research, a decision support system for recommending the selection of maternity hospitals in Pontianak was developed using the Weighted Product (WP) method, with the constructed system in the form of a web-based application. The aim of this study is to facilitate pregnant women in choosing maternity hospitals in Pontianak based on criteria obtained from a survey of pregnant women, including distance, facilities, cost, and reputation. The WP method was applied through three main stages: weight normalization, vector S calculation, and vector V computation for final ranking. Testing in this research involves five alternative maternity hospitals, and each criterion is assessed on indicators ranging from 1 to 5. The results obtained indicate that Anugerah Bunda Khatulistiwa Maternity Hospital achieved the highest final ranking score among all evaluated alternatives. This system is expected to assist expectant mothers in making more informed decisions when selecting a maternity hospital that best suits their needs.

Sentiment Analysis on Electric Vehicles in Indonesia Using Bidirectional Encoder Representations from Transformers (BERT) and Named Entity Recognition (NER) Methods

Billy — 2026-06-15

Air pollution is a major environmental issue due to its significant impact on human health, with the transportation sector being one of the largest contributors. In Indonesia, increasing motor vehicle usage has led to higher greenhouse gas emissions, encouraging the transition toward electric vehicles as a cleaner alternative. However, the adoption of electric vehicles is influenced not only by technical factors such as infrastructure and cost, but also by public perception, which varies across different digital platforms. This study aims to analyze public sentiment toward electric vehicles in Indonesia using a Natural Language Processing (NLP) approach by combining Bidirectional Encoder Representations from Transformers (BERT) and Named Entity Recognition (NER). BERT is utilized to classify sentiments into positive, negative, and neutral categories by considering bidirectional contextual information, while NER is used to identify key entities such as companies, products, locations, and issues discussed in public discourse. The results show that the BERT model achieves an accuracy of 71.05%, precision of 61.31%, recall of 59.28%, and a misclassification error of 28.95%, indicating a fairly good performance in sentiment classification. Furthermore, NER analysis reveals that event and opinion are the most influential factors affecting public interest, followed by company, product, and quality, while location, price, action, and feature have lower influence. Overall, public interest in electric vehicles in Indonesia is relatively high but dynamic, as it is strongly influenced by circulating information and public opinion.

Implementation of the Gradient Boosting Algorithm for Palm Oil Price Prediction

Wilbert Fernando — 2026-06-15

The price of palm oil is highly volatile due to the influence of global market dynamics, trade policies, and climate change, creating uncertainty for industry players in decision-making. This research aims to implement the XGBoost (Extreme Gradient Boosting) algorithm, optimized using Grid Search Cross-Validation, to predict palm oil prices. The dataset used is the Palm Oil Futures Historical Data.csv obtained from Kaggle, consisting of nine features. Data preprocessing is performed using StandardScaler for normalization, followed by model training with hyperparameter tuning. The system is built as a web-based application separating the frontend using PHP and Flask as the Backend API. Testing on 105 test data points yielded an MAE of 43.97, RMSE of 65.14, and R² of 91.82%, demonstrating the model’s strong ability to explain palm oil price variation. Based on the results, the XGBoost algorithm is suitable as a decision-support tool for commodity price prediction, achieving high accuracy consistent with standard criteria for commodity price forecasting and capable of handling large datasets.

Implementation of the Random Forest Algorithm for Loan Eligibility Prediction and Feature Analysis Based on Financial Data

Angel — 2026-06-15

The advancement of information technology has led to an increasing demand for loan access, both through banking institutions and online lending platforms. However, the process of evaluating loan eligibility, which is still carried out manually or semi-manually, is prone to human error and decision-making bias, ultimately increasing the risk of loan defaults. This study aims to implement the Random Forest algorithm to predict loan eligibility based on financial data, as well as to evaluate its accuracy. The dataset used in this study is loan_approval_dataset.csv, which is downloaded from Kaggle, utilizing 11 input features. The system is developed as a web-based application using Laravel as the main frontend and backend framework, while Flask is used as a backend API for executing the machine learning processes. The testing results show that the Random Forest model achieves an accuracy of 98.44%, with a precision of 98.14%, recall of 99.37%, and an F1-score of 98.75%. Furthermore, the cibil score feature is identified as the most influential factor in the prediction process, contributing 80.65% to the model's outcome. These findings indicate that the Random Forest algorithm is highly effective for use in a loan eligibility prediction system, as it provides fast, objective, and highly accurate results.

Influencing the Success of SPBE Jambi Provincial Government Using the SEM Method

Chandy Ophelia S — 2026-06-15

The Electronic-Based Government System (SPBE) is a strategic instrument for digital governance in Indonesia, but its success in local government remains uneven. The problem addressed in this study is the fluctuating SPBE performance of the Jambi Provincial Government, which is associated with network instability, changing application coordinators, overlapping data input, diverse user age groups, and uneven digital literacy among state civil apparatus (ASN) and service users. This study aims to identify the determinants of SPBE success from the ASN perspective and to explain how system quality and information quality shape perceived ease of use, perceived usefulness, user satisfaction, and net benefits. A quantitative explanatory survey was conducted with 385 ASN respondents who interact with SPBE services in the Jambi Provincial Government. The research model integrates constructs from technology acceptance and information system success perspectives and was tested using partial least squares structural equation modeling. The measurement results show that all indicators are valid, with outer loading values above 0.70, AVE values above 0.50, and Cronbach alpha values above 0.80. The structural results indicate that information quality has the strongest effect on perceived usefulness (beta = 0.861), followed by user satisfaction on net benefits (beta = 0.844) and system quality on perceived ease of use (beta = 0.834). Perceived usefulness also has a stronger effect on user satisfaction than perceived ease of use. These findings confirm that SPBE success in Jambi depends primarily on accurate, complete, timely, and relevant information that creates real work benefits and sustained user satisfaction.

Effective Strategies for Memorizing Mathematical Formulas in a Literature Review Study

Feronika Br Siahaan — 2026-06-15

Mathematics is often perceived as a difficult subject due to the large number of formulas that students must understand and memorize. This condition can lead to learning difficulties and trigger mathematics anxiety, which may reduce students’ ability to retain mathematical concepts. This study aims to examine various effective strategies for memorizing mathematical formulas based on previous research. The research employed a qualitative approach using a literature review method involving 30 relevant scientific articles obtained from several academic databases. Data were collected through a literature study, while the analysis was conducted by identifying, comparing, and synthesizing findings from the collected literature. The results show that strategies for memorizing mathematical formulas can be categorized into four main groups: audio-musical and artistic strategies, digital technology and gamification innovations, cognitive strategies through mnemonics and kinesthetic tools, and structured drill methods. These strategies have been shown to improve memory retention, learning motivation, and students’ learning outcomes. The findings indicate that the application of creative and multisensory learning strategies can help students memorize mathematical formulas more effectively.

Analysis of Taxsee Driver User Satisfaction in Jambi City Using the Servqual Method

Josefi Virgi Narada — 2026-06-15

This research analyzes driver satisfaction levels in Jambi City using the Taxsee Driver application through the Service Quality (SERVQUAL) method. The study background is user complaints regarding GPS inaccuracies and the automatic order system that affects driver performance ratings. Data were collected via online questionnaires from 385 active drivers in Jambi City and processed using Structural Equation Model with SmartPLS. Results show that three of five SERVQUAL dimensions significantly affect user satisfaction: Tangibles (T-Statistic 5.073), Responsiveness (T-Statistic 3.782), and Empathy (T-Statistic 4.026). Reliability (T-Statistic 1.735) and Assurance (T-Statistic 1.303) were not significant. The R-Square value of 0.879 indicates the model explains 87.9% of user satisfaction variation. Developers are recommended to improve navigation accuracy and responsiveness to maintain driver partner loyalty.

Implementation of a Web-Based Decision Support System for New Employee Recruitment Using the VIKOR Method

Arochman — 2026-06-15

An effective and objective employee selection process is essential to obtain high-quality human resources. This study aims to develop a web-based decision support system to assist in the recruitment of new employees using the VIKOR method. The VIKOR method is chosen because it can rank alternatives based on their closeness to the ideal solution while considering compromise among criteria. The criteria used in the system include education, work experience, skills, interview results, and work personality. This research adopts the waterfall approach for system development and implements PHP programming language with a MySQL database. The testing results indicate that the system is capable of providing accurate and consistent rankings of job candidates, as well as facilitating the HR team in conducting evaluations more efficiently.

Design of 3D Puzzle Game "Moodoria" Using Unity as an Educational Media for Emotional Intelligence

Bryan Anderson Basli — 2026-06-15

Emotional awareness is crucial for mental health, yet conventional education methods are often less engaging for adolescents. This study aims to design and develop a 3D puzzle game called "Moodoria" using Unity as an interactive medium for emotional intelligence education. The research method used is Research and Development (R&D) with the Game Development Life Cycle (GDLC) model, including needs analysis, literature study, concept design, implementation, and user testing. 3D assets were created using Blender. The game was tested on a small group of users (5–10 people) using a Likert-scale questionnaire. Results show that all main features (menu navigation, character movement, object interaction) function well. User assessments scored high on gameplay (4.87 for challenge) and enjoyment (4.67), and the game was considered feasible as an emotional education medium (average score 4.23). In conclusion, "Moodoria" was successfully developed as an engaging educational game, although sound effects and character expression variations need improvement.

Classification of Handwriting Margin Patterns Using Ensemble Bagging Decision Tree

Rista Ifanka — 2026-06-15

Analysis of handwriting margins plays an important role in graphology, as margin patterns are often associated with individual behavioral tendencies and personality traits. Therefore, detecting and classifying margin characteristics is essential to support automated handwriting analysis using computational approaches. This study uses a computer vision-based approach to classify left-margin handwriting patterns into widening and narrowing categories. The classification is performed by analyzing margin characteristics extracted from scanned handwriting images. The processing pipeline consists of image preprocessing, hybrid feature extraction, and classification using an Ensemble Bagging Decision Tree model. The preprocessing stage enhances image quality through grayscale conversion, contrast adjustment, adaptive thresholding, and noise removal, followed by Region of Interest extraction to focus on the handwriting area. The feature extraction stage applies a hybrid strategy that combines line-based margin analysis and global spatial features to capture both local variations and overall structure. Model performance is evaluated using stratified 5-fold cross-validation to ensure reliable and unbiased results. The experimental findings show that the method achieves an average accuracy of 84.91%, with relatively balanced precision, recall, and F1-score across both classes. These results indicate that margin-based features are effective for representing handwriting patterns in classification tasks. However, variations in writing style and noise from the scanning process may influence performance. Overall, this study demonstrates that the applied approach provides reliable classification results and has potential for further improvement through feature expansion and more advanced learning models.

Sentiment Analysis of SPayLater and SPinjam Features in the Shopee Application Using the Support Vector Machine (SVM) Algorithm

Rahmad Rahmad Nawi Pane — 2026-06-15

The rapid development of information technology and the increasing use of e-commerce applications have generated a large number of user reviews that can be used to measure user satisfaction. SPayLater and SPinjam, as features in the Shopee application, receive various responses in the form of positive, negative, and neutral sentiments, making automatic sentiment analysis necessary. This study aims to analyze user sentiment and implement the Support Vector Machine (SVM) algorithm to classify reviews. The data used consist of 500 user reviews obtained from the Google Play Store. The method includes preprocessing, labeling, and classification using SVM. The results show that there are 231 positive, 230 negative, and 39 neutral sentiments. Model evaluation yields an accuracy of 74%, precision of 0.78, and recall of 0.84, indicating that the model performs fairly well. The developed system is also capable of processing data automatically and displaying classification results effectively. Therefore, the SVM algorithm is effective for sentiment analysis of SPayLater and SPinjam services in the Shopee application.

Analysis of Green Computing Implementation Strategies for Energy Efficiency in Server Infrastructure

Daniel Rionaldo — 2026-06-15

The rapid development of information technology has increased the use of server infrastructure in various organizations and data centers, resulting in higher energy consumption and operational costs. Green computing is considered an effective approach to improve energy efficiency while reducing environmental impacts. This study aims to analyze green computing implementation strategies for improving energy efficiency in server infrastructure. The research used a descriptive qualitative method through literature studies and comparative analysis of energy management strategies in data centers. The strategies analyzed include server virtualization, server consolidation, energy-efficient hardware, and cooling system optimization. The results indicate that the implementation of green computing can significantly reduce energy consumption compared to conventional server systems. In addition, the implementation improves operational efficiency, reduces electricity usage, and supports environmentally sustainable data center management. Therefore, green computing can be considered an effective solution for developing efficient and environmentally friendly server infrastructure.

Implementation of Random Forest Algorithm for Classifying Land and Building Tax Arrears and Risk Factor Analysis Dashboard

Risky Firmansyah Manik — 2026-06-15

This study aims to develop a predictive model to identify the potential for land and building tax arrears and analyze the dominant risk factors contributing to non-compliance. The research utilizes the Random Forest classification algorithm applied to historical tax data from the Regional Financial and Revenue Management Agency of Binjai City. The approach involves data preprocessing, feature engineering including target encoding for geographical areas, and model training with hyperparameter tuning to optimize classification performance. Furthermore, a web-based interactive dashboard is developed using the Flask framework to visualize the predictions and risk factors. The results demonstrate that the Random Forest model achieves a robust and consistent accuracy of approximately 85% in classifying compliant and non-compliant taxpayers. Feature importance analysis reveals that land area is the most dominant risk factor influencing tax arrears, significantly outweighing other variables. In conclusion, the integration of the Random Forest algorithm with an interactive dashboard provides a highly accurate, efficient, and scalable solution for local governments to transition from reactive tax collection to proactive, data-driven risk management.

Design of a Multi-Tenant Waste Management System with Volume Estimation and Vehicle Trip Optimazation

Intan Nur Sifa — 2026-06-15

Waste management at the village level still faces a number of challenges, such as unstructured waste volume recording, suboptimal collection scheduling, and a lack of transparency in cost management. This study aims to design a multi-user waste management system equipped with a volume estimation model and vehicle route optimization. The approach applied includes a literature review to analyze system requirements, followed by design using flowcharts, Data Flow Diagrams (DFDs), and Entity-Relationship Diagrams (ERDs). The research findings indicate that the developed system successfully integrates the management of waste source data, transportation processes, and cost calculations in a structured manner. The volume estimation model is used to estimate the amount of waste in the field, while route optimization determines the number of vehicle trips based on their carrying capacity. Additionally, the multi-tenant concept allows this system to be used by various regions simultaneously while ensuring data separation. Therefore, this system is expected to improve operational efficiency, management transparency, and the quality of waste transportation services.

Application of Data Mining using the Apriori Algorithm in Analyzing Subject Selection Patterns of Tutoring Students

Rizky Ferdiansyah — 2026-06-15

This study examines the application of data mining using the Apriori algorithm to analyze subject selection patterns among tutoring students in Kudus, Central Java. With the increasing number of students attending tutoring, understanding subject selection patterns is crucial to improve the effectiveness of educational services. The Apriori algorithm, a popular association rule mining technique, is used to identify relationships between frequently selected subjects. The research dataset consists of student subject selection transaction data, including information such as student name, student ID number, tutoring branch, and selected subjects. The analysis process included data preprocessing, data transformation into transaction format using Transaction Encoder, application of the Apriori algorithm with a minimum support of 0.05, and formation of association rules with a minimum confidence of 0.3. The results show frequent itemsets indicating the most popular subjects and association rules that describe students tendencies in selecting subject combinations. These findings can be utilized by tutoring managers to design more effective learning packages, optimize the allocation of teaching resources, and provide subject recommendations tailored to student needs. This research contributes to the development of educational data mining in the context of tutoring institutions in Indonesia.

Comparative Analysis of K-Means Clustering and K-Medoids Clustering Methods in Clustering Neonatal Infant Mortality Rates in West Java Province

Intan Putri Septiyani — 2026-06-15

Neonatal mortality rate is an important indicator in assessing public health conditions. This study aims to cluster neonatal mortality data in West Java Province using the K-Means Clustering and K-Medoids Clustering methods, as well as compare the performance of both methods in producing the best clusters. The study used secondary data obtained from Open Data West Java. The research stages included data selection, preprocessing, clustering, and evaluation using the Davies-Bouldin Index (DBI). The experiments were conducted using cluster variations (k) from 2 to 8. The results showed that the K-Means Clustering method produced the best performance with a DBI value of 0.430 at k = 3. The clustering results generated three categories: low-risk cluster with 408 data points, medium-risk cluster with 65 data points, and high-risk cluster with 13 data points. The differences in cluster characteristics indicate variations in neonatal mortality risk levels among regions in West Java Province. The findings of this study are expected to support decision-making and more targeted health policy planning.

Keywords: K-Means Clustering, K-Medoids Clustering, Davies-Bouldin Index, Neonatal Mortality.

Analysis and Design of the Nusa Graha Module for Village Asset Management and Facility Booking on the NUSAEKA Multi-Tenant SaaS Platform

Purnia Setiawati — 2026-06-15

In most regions of Indonesia, village asset management and the process of booking village facilities are still carried out manually, which can lead to disorganized record-keeping, data loss, and a lack of access for village residents. This study was conducted to analyze and evaluate the Nusa Graha module as a component of the Nusaeka multi-tenant SaaS platform, focusing on village inventory management, automatic asset depreciation, and web-based village booking services. This research was conductes through a literature review and system analysis obtained through consultation with supervising lecturer as well as document analysis. The analysis results include business flowcharts, Data Flow Diagrams (DFDs) at levels 0 and 1, and Entity-Relationship Diagrams (ERDs), which consist of several main tables. The research findings indicate that the Nusa Graha module can support and streamline asset management and the structured process of facility rentals using multi-tenant data via tenant_id and a modular language. Additionally, the Nusa Graha module facilitates integration with the Nusa Artha financial module if the village subscribes to it.

Customers’ Loss of Confidence in Banking Security Systems: A Case Study of the Loss of BRI Customers’ Funds

Aisyah Safitri — 2026-06-15

The phenomenon of customer funds going missing in the banking sector, particularly in the case of Bank Rakyat Indonesia (BRI), has raised concerns about the security of the digital banking system and has led to a decline in public confidence. This study aims to analyse the crisis of customer confidence in banking security systems by examining influencing factors, such as cyber risk, risk perception, and the role of social media. The research method employed is a qualitative approach using case studies, utilising secondary data obtained from academic journals, institutional reports, and case documentation.
The research findings indicate that the loss of customer funds is influenced by vulnerabilities in digital security systems and the rise in cybercrime, such as phishing and social engineering. Furthermore, these incidents have led to a decline in customer trust, a trend exacerbated by the dissemination of information via social media.
This study concludes that the crisis of customer trust is caused not only by technical factors, but also by risk perceptions and the dynamics of public information. Therefore, improvements in banking system security, strengthened consumer protection, and effective communication strategies are required to maintain customer trust.

Design of a Web Based Population Data Information System at Matawai Atu Village Office

Jesika Prince Piri — 2026-06-15

The development of information technology has greatly influenced many sectors, including village administration. The Matawai Atu Village Office, located in Umalulu Subdistrict, East Sumba Regency, still uses a manual system to record population data such as births, deaths, new residents, and relocations. Data is recorded in a main register book and then processed using Microsoft Word to create reports. This method causes several problems, including the risk of data loss, data entry errors, slow data searching, and delays in report preparation. To solve these problems, this study aims to design a web-based population data information system that is effective and efficient. The study uses the Waterfall method, which includes the stages of requirements analysis, system design, implementation, testing, and maintenance. The system is developed using PHP and a MySQL database. Data collection is carried out through interviews, direct observation at the research location, and literature study. System testing is conducted using Black Box Testing to ensure that all features work properly, and the System Usability Scale (SUS) to measure how easy the system is for users. The results show that the developed system can manage population data more accurately, quickly, and securely. The system also makes it easier for staff to search data, manage documents, and prepare reports. With this system, it is expected that public services at the Matawai Atu Village Office will improve and better support the work of village staff.

Implementation of a Chatbot-Based AI Agent for Employee and Student Attendance Systems with Face Recognition and N8N Integration

Muh. Dwicky P. Sanjaya — 2026-06-15

Students frequently rely on direct messaging to verify the presence of lecturers and staff on campus, a practice that often results in delayed responses due to the recipients' busy schedules. This study aims to design, implement, and evaluate an automated attendance system based on an AI agent utilizing face recognition technology and n8n as a centralized workflow automation platform. The research employs a Research and Development (R&D) approach with the Agile development method. Real-time face detection and recognition are performed from CCTV camera feeds using a Python module that integrates the InsightFace and MediaPipe algorithms. Identified attendance data is automatically stored in Google Sheets, subsequently processed by n8n to deliver information to users via a WhatsApp chatbot powered by the Gemini 2.5 Flash model. Testing conducted on 419 samples yielded an accuracy of 86.16%, with 275 True Negative values demonstrating the system's capability in filtering unregistered faces. The overall average system latency was 15.9 seconds, with a chatbot automation response time of only 9.3 seconds. This research demonstrates that the integration of workflow automation and AI agents is effective in improving the efficiency of academic attendance information access.

Selection of Outstanding Lecturers Using the Simple Multi-Attribute Rating Technique (SMART) Method

Dede Irmayanti — 2026-06-15

Lecturers play a crucial role as professional educators in the implementation of higher education through the Tridarma Perguruan Tinggi (Triple Dharma of Higher Education), which encompasses education, research, and community service. The selection of exemplary lecturers serves as both a form of recognition and a motivational instrument to enhance institutional quality. However, the selection process is often hindered by subjective assessments and the lack of standardized measurement, which may lead to dissatisfaction and diminish the objectivity of the results. This study aims to address these issues by implementing a Decision Support System (DSS) using the Simple Multi-Attribute Rating Technique (SMART) method. The SMART method was selected for its effectiveness in facilitating multi-criteria decision-making through weight assignment to priority parameters, such as scientific publications, educational qualifications, and external achievements. The results of this system implementation are provide structured, transparent, and accurate decision recommendations, ensuring that the selection of exemplary lecturers is conducted objectively based on measurable data.

Design of an Android-Based Sitting Posture Detection Application Using Deep Learning

Jhonshen Lim — 2026-06-15

Prolonged poor sitting posture is a major cause of musculoskeletal disorders including lower back pain and spinal abnormalities. This study designs and implements PosturApp, a deep learning-based Android application for real-time sitting posture detection using Kotlin. A Multi-Layer Perceptron (MLP) model was trained on 3,526 keypoint datasets sourced from the Kaggle public dataset (Posture Recognition) and direct image capture using an Android front camera, extracting 66 coordinate values from 33 body landmarks via MediaPipe BlazePose. The model was converted to TensorFlow Lite (TFLite) format at approximately 78 KB for on-device inference without internet connectivity. Evaluation results show an accuracy of 97.81% with precision 0.99, recall 0.99, and F1-Score 0.98. The application provides real-time visual feedback through interface color changes and corrective notifications, along with a gallery-based classification feature. Functional testing across eight posture scenarios yielded entirely correct results with confidence values ranging from 59% to 99%.

Design of a Warehouse Inventory Management System Using FEFO Method in NUSA Niaga Multi-Tenant

Lina Nur Afifah — 2026-06-15

Most village-based businesses still manage their inventory manually, from monitoring warehouse stock and generating reports and checking items based on expiration dates. This process is considered inefficient and carries the risk of errors in data recording and reporting. In the process of transferring inventory to the display on the web-based NUSA Niaga platform, the FEFO method is applied. Needs analysis, system design, design implementation, and system documentation were conducted. Literature review on management systems, the FEFO method, multi-tenant architecture, and RBAC were used for data collection. The system was designed to monitor inventory, manage products nearing expiration, record goods transfers, and implement multi-tenant functionality using flowcharts, ERDs, and DFDs. This system is expected to help manage BUMDes warehouse in a more connected and structured manner.

Analysis of JKN Mobile User Satisfaction using SVM and KNN Methods Through PSO Optimization

Esty Purwaningsih — 2026-06-15

This study was conducted to evaluate the service quality of the JKN Mobile application developed by the Health Social Security Administering Agency (BPJS Kesehatan) as a means of facilitating participants in accessing health services. Although the application provides convenience for users, there are still various complaints indicating that the service is not running optimally. Therefore, this study aims to analyze the positive and negative sentiments of JKN Mobile application users by comparing the Support Vector Machine (SVM) and K-Nearest Neighbor (KNN) algorithms based on Particle Swarm Optimization (PSO). The research method was carried out by processing user review data using sentiment classification techniques. The test results showed that the SVM algorithm obtained an accuracy of 85.02% with an AUC value of 0.815, while the PSO-based SVM increased to 86.71% with an AUC of 0.831. The KNN algorithm obtained an accuracy of 39.54% with an AUC of 0.500, while the PSO-based KNN increased to 87.05% with an AUC of 0.736. The results of the study prove that the implementation of PSO is able to improve the accuracy performance of both algorithms.

The Effect of Mobile Banking Usage on Banking Customer Satisfaction

Ayu Maulidia — 2026-06-15

The rapid development of information technology has encouraged the banking sector to innovate through digital-based services, one of which is mobile banking. This service provides convenience for customers in conducting banking transactions quickly, effectively, and efficiently through smartphones without having to visit bank offices. This study aims to determine the effect of mobile banking usage on customer satisfaction in banking services. The research method used is a quantitative descriptive approach with data obtained from previous journals and supporting literature related to mobile banking and customer satisfaction. The results indicate that the use of mobile banking has a positive and significant effect on customer satisfaction. Factors such as ease of use, transaction speed, security, service quality, trust, and digital banking innovation are proven to influence customer satisfaction in using mobile banking services. In addition, digital banking services are able to improve customer convenience and efficiency in carrying out financial transactions. Therefore, banks are expected to continuously improve the quality of digital services, strengthen transaction security systems, and provide sustainable innovations to maintain customer satisfaction and loyalty in the digital era.

Evolution and Impacts of AI-Based Rainfall Prediction Systems on Agricultural Management in Tropical Regions: A 20-Year Systematic Review

Safrizal — 2026-06-15

Global climate change has significantly disrupted rainfall patterns in tropical regions, posing major challenges to agricultural productivity and food security. Accurate rainfall prediction has become a critical component of data-driven agricultural management. This study conducts a systematic literature review (SLR) following the PRISMA 2020 guidelines to analyze the evolution of AI-based rainfall prediction systems and their multidimensional impacts on tropical agricultural management over the period 2008–2026. Data were sourced from Scopus using three Boolean search strings, yielding 239 records, of which 235 articles were retained after duplicate removal and quality assessment using the Mixed Methods Appraisal Tool (MMAT) with a threshold score of ≥5. Bibliometric analysis was conducted using VOSviewer and Bibliometrix (R), while thematic narrative synthesis was performed using NVivo 14. Results reveal a clear four-phase technological evolution: conventional methods (2008–2015), machine learning adoption (2016–2020), deep learning and IoT integration (2021–2023), and multimodal and large language model era (2024–2026). Technical impacts dominated the corpus (accuracy improvements of 18–35%), while social and economic impact studies remain critically underrepresented (2.6% and 0.9%, respectively). Key research gaps identified include poor model interpretability (black-box problem), limited integration with decision support systems (DSS), inadequate tropical-specific model development, and the near-total absence of longitudinal impact evaluations. This study contributes a holistic synthesis integrating technological evolution with multidimensional impact analysis, offering strategic recommendations for developing more adaptive, transparent, and equitable AI rainfall prediction systems aligned with SDG 2, SDG 13, and SDG 15

Performance Evaluation of the BERT Model in Sentiment Analysis of DANA Application User Reviews

Hazael Susanto — 2026-06-15

The rapid growth of digital wallets in Indonesia generates a large volume of user reviews on platforms such as the Google Play Store that cannot be efficiently analyzed manually. This study aims to evaluate the performance of the BERT (Bidirectional Encoder Representations from Transformers) model in sentiment classification tasks on a dataset of DANA application user reviews collected from the Google Play Store. The BERT model is fine-tuned using labeled Indonesian-language data with three sentiment classes: positive, negative, and neutral. Specialized preprocessing strategies are applied to handle the characteristics of informal text, abbreviations, and code-switching phenomena prevalent in Indonesian user reviews. Evaluation is conducted using accuracy, precision, recall, and F1-score metrics. Experimental results indicate that the fine-tuned IndoBERT model achieves an accuracy of 91.24% with a weighted F1-score of 0.91 on a test dataset of 6,106 samples. The Negative class achieves the highest performance with an F1-score of 0.95, followed by the Positive class (0.88) and Neutral class (0.84). This study provides empirical evidence of the effectiveness of the IndoBERT Transformer architecture for sentiment analysis in the Indonesian-language fintech domain and can serve as a reference for developing deep learning-based NLP systems in similar contexts.

Design of a Web-Based Village Tourism Management Information System with Multi-Tenant Architecture as an Integrated Platform: The Nusa Wisata Module

Mayza Nurul Khasanatun Nisa mayza — 2026-06-15

The rapid development of information technology has significantly impacted various sectors, including tourism. However, in practice, village tourism management is still commonly conducted conventionally and lacks integration, resulting in various issues related to service efficiency, data management, and operational transparency. This study aims to design a web-based village tourism management information system using a multi-tenant architecture approach in the Nusa Wisata module as part of the integrated Nusa Eka platform.The research method employed in this study is system design using the System Development Life Cycle (SDLC) approach, focusing on the requirements analysis and system design stages. System modeling was conducted using Entity Relationship Diagram (ERD) for database design and flowcharts to illustrate the system process flow. The results of this study are in the form of a system design capable of integrating the management of multiple villages within a single platform while maintaining data separation through the use of the tenant_id attribute. The system is designed with several main features, including e-ticketing, tourism package and attraction management, tourism operational staff management, and parking management with an automated revenue-sharing mechanism. In addition, the system supports integration with other modules within the Nusa Eka platform, such as Nusa Praja, Nusa Graha, and Nusa Artha. Based on the design and analysis results, the proposed system provides a more integrated, flexible, and scalable solution compared to previous systems that still use a single-tenant approach. This system design is expected to improve operational efficiency, data transparency, and service quality in digital-based village tourism management.

Web-Based Congregation Data Management Information System for the Pamalar Sumba Christian Church

Marthen Umbu Delu Palabu — 2026-06-15

Information technology plays an important role in improving efficiency, accuracy, and accessibility in administrative processes. Churches with large and dispersed congregations often face difficulties in managing and searching congregation data. The Sumba Christian Church, located in Umbu Langang Village, Central Sumba Regency, also experiences these challenges. The congregation consists of 1,231 members, including 589 males and 642 females, and the number continues to grow. Currently, congregation data is still recorded manually in books, making data updates slow, inefficient, and prone to errors or data loss due to non-centralized storage. To address these problems, this study developed a web-based Congregation Data Management Information System using the Waterfall development method. The system aims to simplify data recording, updating, and searching processes, making church administration more effective and efficient. The implementation of this system is expected to improve the speed, accuracy, and reliability of congregation data management. System testing was conducted using the Black Box Testing method, which showed that all system features functioned successfully with a 100% success rate. In addition, usability evaluation using the System Usability Scale (SUS) produced an average score of 80.75, indicating that the system is highly usable and acceptable for church administration activities.

Design and Development of a News Website and CMS Based on Three-Tier Architecture: A Case Study of PLTU Pangkalan Susu

Zainuddin — 2026-06-15

Pangkalan Susu Steam Power Plant does not yet have a structured official publication platform, causing information dissemination to depend mainly on social media and making news archives difficult to organize and retrieve regularly. This condition highlights the need for a web-based information system that supports more professional content management. This study aims to design and develop a news website and Content Management System (CMS) by applying a Research and Development (R&D) approach and three-tier architecture, which separates the presentation layer, application logic layer, and data layer. The development process consists of requirements analysis, system architecture and database design, implementation, and functional testing. The system was implemented using Laravel as the application framework, Blade for the user interface, Filament as the administration panel, and MySQL as the database management system. Functional testing was conducted using the black-box method on 14 main scenarios, including the management of news, categories, authors, banners, and logout features. The results show a 100% success rate, indicating that all tested functions operated according to the expected outcomes without functional errors. The proposed architecture produces a modular, scalable, and maintainable system that can support the future development of the official information platform of Pangkalan Susu Steam Power Plant.

Designing the Foundation of a Multi-Tenant and Surrogate Key-Based Nusa Praja Village Government System on the Nusa Eka Platform

Sri Rahayu — 2026-06-15

Village government administration in Indonesia is still largely manual, resulting in service inefficiencies and the potential loss of sensitive population data. The purpose of this research is to create the architectural foundation of a village government information system called NUSA PRAJA, which is part of the NUSAEKA multi-tenant Software as a Service (SaaS) platform. This research applies the Multi-Tenant Isolation concept to ensure data security and separation between villages, as well as the Surrogate Key concept to protect residents' National Identification Number (NIK) data from leaks. The research method used is Waterfall with the stages of requirements analysis, system architecture design, and database design. The results of the research are a multi-tenant system architecture design based on a shared database with tenant_id filters, an Entity Relationship Diagram (ERD) design with 14 main tables, and a flowchart design for three user roles. This project is expected to be a strong foundation for building a secure, scalable, and easy-to-use digital village system for village governments throughout Indonesia.

Designing a Key Performance Indicator Application for Sales Performance Evaluation Using the Web-Based ROC Method (Case Study of PT. Valve Automation Indonesia)

Adam Panji Maulana — 2026-06-15

Sales performance assessment is an important aspect in human resource management because it directly influences the achievement of company targets. At PT. Valve Automation Indonesia, KPI weighting is determined by managers based on personal assessments, this causes weight imbalances when adding or reducing assessment parameters. This study aims to design and implement a web-based Key Performance Indicator (KPI) application with the Rank Order Centroid (ROC) method to determine the weight of criteria based on the order of importance so as to produce a more objective and proportional weighting. The assessment criteria used include Absence, Sales Revenue, New Customers, Payment Collection, Visit Customers, and Follow Up Progress. This application was developed through the stages of needs analysis, design, implementation, and testing. This application shows that it is able to calculate KPI values automatically, generate sales performance rankings, and assist managers in making decisions more effectively and efficiently.

Network Device Performance Monitoring Using the Simple Network Management Protocol (SNMP) Method

Aldi Mulia Rismanto — 2026-06-15

Network problems frequently occur at Politeknik Negeri Pontianak due to the increasing number and scale of network devices. These issues require continuous monitoring to ensure service availability across all network devices. To address this problem, the author conducted network monitoring using the SNMP (Simple Network Management Protocol) method and network performance measurement using the Wireshark application. SNMP is a standard protocol used to monitor and manage network devices such as routers, switches, servers, and other networking equipment. The research stages began with data collection, followed by monitoring and performance testing of the network. After testing the network in the Informatics Engineering Building, both satisfactory and unsatisfactory results were obtained. The results of SNMP measurements on MRTG showed the lowest throughput values on the second day of testing, with 485.6 kbps for daily traffic, 236.8 kbps for weekly traffic, 232 kbps for monthly traffic, and 121.6 kbps for yearly traffic. Meanwhile, the Quality of Service measurement produced the lowest throughput value of 0.225 kbps, packet loss of 0.354%, delay of 3.331 ms, and jitter of 8.763 ms.

Public Sentiment Analysis on the Issuance of Panda Bonds as an Effort for Rupiah Stability using SVM Algorithm on Youtube Social Media

Junjung Rahmat Santosa — 2026-06-15

The stability of the Rupiah exchange rate is a crucial indicator of Indonesia's economic health, one of which is pursued through the issuance of Panda Bonds. However, this policy has triggered dynamic discourse on social media, particularly YouTube. This study aims to map public perception and test the performance of the Support Vector Machine (SVM) algorithm in classifying sentiments related to this issue. The research methodology includes scraping YouTube comment data, text preprocessing, automated labeling using the Lexicon-based method, and classification using SVM with a Linear kernel. From a total of 659 collected data, the results show that public sentiment is dominated by positive responses at 51.9%, followed by neutral sentiment at 29.0%, and negative sentiment at 19.1%. While public concerns focus on the debt burden and foreign currency dependence, there is overall support for economic stability efforts. The model evaluation demonstrates excellent performance, achieving an accuracy rate of 87.86%, precision of 88.79%, and an F1-score of 87.96%. This proves that a hybrid approach between Lexicon-based and SVM is effective in analyzing complex public opinions within the economic domain on social media.

Design of a Multi-Tenant SaaS-Based Centralized Financial System Using a Silent Accounting Approach

Rizki Cahya Putri cahyaputri — 2026-06-15

Village financial management faces various fundamental challenges, including transaction recording that is still manual, a lack of integration between financial systems and village operational activities, and the absence of a platform capable of serving multiple villages within a single efficient infrastructure. These conditions result in financial reporting processes that are inefficient, error-prone, and difficult to account for. This study aims to design a centralized financial system based on a multi-tenant Software as a Service (SaaS) architecture using the Silent Accounting approach, defined as an automated transaction recording mechanism triggered by operational module activities without manual intervention. This study employs a qualitative descriptive method with a literature review approach. The design yields three main artifacts a system flowchart illustrating the workflow from user authentication, role assignment, and transaction validation through to automatic journal entry and posting to general ledger an Entity Relationship Diagram (ERD) modeling the database structure consisting of seven entities and a Data Flow Diagram (DFD) breaking down the system into five main processes A multi-tenant architecture with a ‘tenant_id’ column ensures data isolation between villages while allowing a single platform to serve multiple village simultaneously. The Silent Accounting mechanism ensures that all village financial activities are recorded consistently, accurately and in real time. The design is expected to serve as the foundation for the development and scalable village financial management platform.

Analysis of Green Computing Implementation in Efforts to Improve Resource Efficiency in the Campus Environment

Alfin Budiman Sihotang — 2026-06-15

This study aims to analyze the level of green computing implementation in efforts to improve resource efficiency in the campus environment. The primary problem addressed is the high energy consumption in higher education environments due to the increasing use of information technology devices, necessitating efficient and sustainable energy management measures. This study employs a mixed methods approach, combining literature review with quantitative data collection through questionnaires to obtain data on user understanding and behavior regarding green computing. The results indicate that the majority of respondents demonstrate a good understanding of energy efficiency. Based on the data obtained, the authors conclude that the level of awareness and implementation of green computing among students is very good. The findings also reveal that students have a high concern for the impact of energy consumption on the environment and support energy conservation efforts. Overall, this study demonstrates that the application of green computing has great potential for development through broader research and targeted campus policies. This research is expected to serve as a foundation for developing more efficient energy policies in higher education.

Performance Evaluation of Machine Learning Algorithms in Sentiment Analysis of Spotify Reviews

Frizi Olivian — 2026-06-15

The rapid growth of digital music streaming platforms has generated a massive volume of user reviews on the Google Play Store, making manual analysis practically infeasible. This study evaluates and compares the performance of three machine learning algorithms Support Vector Machine (SVM), Neural Network (Multilayer Perceptron), and Random Forest in classifying sentiments from Spotify user reviews written in Indonesian. A total of 10,000 reviews were collected from the Google Play Store using the google-play-scraper library and processed through a text preprocessing pipeline comprising cleaning, case folding, word normalization, tokenization, stopword removal, and stemming using the Sastrawi library. Sentiment labeling was performed automatically using the InSet lexicon, categorizing reviews into three classes: Positive (56.63%), Neutral (30.60%), and Negative (12.76%). Feature extraction was conducted using the TF-IDF method, with an 80:20 train-test split strategy and stratified sampling to maintain class distribution. Model performance was evaluated based on accuracy, precision, recall, and F1-score metrics. The results demonstrate that SVM and Neural Network achieved equivalent and superior accuracy of 0.937, with macro F1-scores of 0.908 and 0.907, respectively, outperforming Random Forest which recorded an accuracy of 0.853 and a macro F1-score of 0.777. These findings indicate that SVM and Neural Network are more optimal and reliable for sentiment classification of Indonesian-language Spotify reviews, while Random Forest requires further improvement, particularly in recognizing minority classes.

Topic Modeling of Clash of Clans Player Reviews Using NLP-Based Latent Dirichlet Allocation (LDA) Machine Learning Method

Rai Markus Panamuan — 2026-06-15

The rapid growth of the mobile gaming industry has generated millions of player reviews on platforms like the Google Play Store. Clash of Clans, developed by Supercell, is one of the world's most popular mobile strategy games, generating a vast volume of user reviews that are difficult to analyze manually. This study applies Latent Dirichlet Allocation (LDA), a generative probabilistic machine learning model based on Natural Language Processing (NLP), to identify and cluster key topics discussed in player reviews on the Google Play Store. A total of 10,000 player reviews were collected through web scraping, followed by NLP-based text preprocessing including tokenization, stopword removal, and lemmatization. The LDA model was optimized using a coherence score evaluation of 0.512, resulting in the identification of five dominant discussion topics: technical issues and bugs, game updates and balance, gameplay and strategy, monetization and in-app purchases, and social interactions and clan systems. The results show that LDA-based topic modeling provides structured and actionable insights for game developers to understand player feedback and improve game quality. This research contributes to the field of NLP-based mobile game review analysis.

Integration of the Webqual 4.0 and UEQ Methods in Evaluating the Impact of the Student Recruitment Information System on Business Efficiency and Admissions Services (Case Study: Sari Mutiara Indonesia University)

Lius Luaha — 2026-06-15

This study aims to evaluate the service quality of the New Student Admission Information System (PMB) at Sari Mutiara Indonesia University and its impact on business efficiency using the WebQual 4.0 method and the User Experience Questionnaire (UEQ). Based on a population of 1,700 users, the sample size was determined using the Slovin formula (5% margin of error), resulting in 324 respondents (319 prospective students, 5 admissions staff) selected via purposive sampling. The WebQual 4.0 evaluation yielded an average score of 4.05 (Good category), with the Usability dimension scoring highest (4.30) and Service Interaction lowest (3.80). UEQ analysis showed that all scales fell within the positive category (>0.8), with the main strength lying in the pragmatic aspect of Perspicuity (2.36), whilst recording the lowest score in the hedonic aspect of Novelty (1.47). Operationally, the system’s high ease of use has been shown to reduce data verification time by 75% (from 20 to 5 minutes) and lower the human error rate from 12% to below 2%. In conclusion, the PMB system operates very efficiently from a business perspective and is user-friendly; however, an update to the user interface (UI/UX) is urgently needed to enhance the novelty value, along with the integration of a more responsive technical support service.

Implementation of the K-Means Method for Developing an Air Quality Monitoring Information

Faisal Rifky Nugraha — 2026-06-15

Air quality in Bogor City is becoming increasingly complex due to the rising number of motor vehicles, small-scale industrial activities, and seasonal dynamics that are difficult to analyze using conventional methods. The Environmental Agency of Bogor City routinely monitors air quality through the Air Quality Monitoring System (AQMS); however, data utilization remains confined to monitoring and reporting, necessitating advanced analysis to achieve a more comprehensive view of air quality patterns. This study aims to classify time periods based on air quality parameters using the K-Means clustering method to identify good and bad air pollution categories. The research data was obtained from the Bogor City Environmental Agency's AQMS for the period from January 2023 to June 2025. The results indicate that the K-Means method successfully clustered the data into two groups: good and bad air quality categories. Good air quality was identified in 2023 (January, February, March, April, November, and December), 2024 (January, February, March, April, October, November, and December), and 2025 (January to May). Conversely, poor air quality occurred in 2023 (May to October), 2024 (May to September), and 2025 (June). The findings of this research are expected to support pollution control strategies and early warning systems based on air quality data.

Analysis of Trends and Development of Low-Light Image Enhancement Methods in Computer Vision

Ani Sanirah — 2026-06-15

This study investigates the development of Low-Light Image Enhancement (LLIE) methods in the field of computer vision using a Systematic Literature Review (SLR) approach. The review was conducted on 56 scientific articles selected from a total of 604 papers entirely sourced from the Scopus database based on the PRISMA 2020 guidelines. The results indicate that LLIE research has evolved from traditional methods, such as histogram equalization and Retinex, toward deep learning-based approaches including CNN, GAN, Transformer, and diffusion models. Modern methods have demonstrated superior performance in improving image illumination, preserving details, and reducing noise. In addition, real-world datasets and zero-reference approaches are increasingly adopted to improve model generalization capability. However, challenges remain regarding computational complexity, detail preservation, and model performance under extreme low-light conditions. This study concludes that future LLIE research will focus on developing models that are more adaptive, efficient, lightweight, and robust for various computer vision applications.

Analysis and Simulation of a Queueing System in a Self-Service Seblak MSME using the FIFO Model

Hayati — 2026-06-15

Micro, Small, and Medium Enterprises (MSME) Seblak Buffet often experience long queues and inefficient waiting times, especially during peak hours, which negatively impact customer comfort and service quality. This study aims to analyze and analyze the business's queue system using the FIFO (First In First Out) model to improve service efficiency and fairness. Using a quantitative approach with M/M/1 modeling, on arrival times and following exponential distribution service, data were obtained from direct observation. The simulation results show that although FIFO maintains order, waiting times increase significantly during peak hours due to limited facilities. Therefore, this study recommends adding waiters or rearranging service flows during peak hours as an applicable solution to improve quality and customer satisfaction.

Development and Evaluation of a Desktop Academic Data Encryption Application Using AES with Password-Based Key Management

Muhammad Wahyu Rizqi Pratama — 2026-06-15

This study evaluates a desktop-based academic data security application integrating the Advanced Encryption Standard (AES) and Password-Based Key Derivation Function 2 (PBKDF2) at Pondok Pesantren Nur Rohman. Using an experimental approach, three AES key lengths (128-bit, 192-bit, and 256-bit) were tested against 10 authentic academic files to assess computational performance and cryptographic strength. The results show that AES-128 achieves the fastest speed with a throughput of 57.84–65.00 MB/s, while AES-256 requires the longest processing time due to its 14 internal rounds. Regarding security, all variations exhibit ideal mathematical resilience: ciphertext entropy ranges tightly between 7.9988 and 7.9999 (near absolute randomness), and avalanche effect percentages remain stable between 49.85% and 50.19%. Furthermore, PBKDF2 successfully mitigates brute-force vulnerabilities by mapping passwords into precise hexadecimal keys. The process introduces a minimal, constant file size overhead of exactly 64 bytes (Salt, IV, and padding). Hardware utilization is exceptionally efficient, recording low CPU usage (0.1%–0.2%) and stable RAM allocation (53%–54%). In conclusion, the system delivers an optimal equilibrium between high-level data protection and local hardware efficiency.