Journal of Computer Science and Informatics Engineering (J-Cosine)

Development of a Web-Based System for Scientific Publication Data Management Using NoSQL Architecture

Luthfiyyah Az Zahro — 2025-12-31

This study developed a web-based system for managing scientific publication data using a NoSQL architecture (MongoDB) and an Extract, Load, Transform (ELT) approach. The website allows users to easily access, manage, and visualize data through an interactive interface built with React.js on the frontend and Express.js on the backend. The system supports data input via CSV uploads and manual form submissions, with all data stored in MongoDB Atlas. Data transformation is handled using the Aggregation Pipeline, which performs normalization, validation, and upsert operations to ensure consistency and prevent duplication. Functional testing was conducted on 503 real publication records, while performance testing used 1,000 dummy records. The results show that the system can support up to 20 concurrent users with an average response time of 4,792 ms and no errors. Overall, the website offers a flexible and integrated solution for managing scientific publications in higher education institutions.

Integrated Kinematic-Dynamic Modeling and Ontology-Based Design of a Two-Link Planar Robotic Manipulator

Giri Wahyu Wiriasto — 2025-12-31

Implementation of Profile Matching Method in Student Major Information System at SMAN 4 Mataram

Nazmi Wardiani — 2025-12-31

Education is a crucial component in a country's development, and the quality of education is a key factor in determining national progress. In Indonesia, the Ministry of Education, Culture, Research, and Technology has implemented the Merdeka Curriculum, which provides students with greater flexibility to choose subjects based on their interests and talents. At SMAN 4 Mataram, the subject selection process starting in the second semester of grade 10 still faces challenges, such as students' difficulty in determining suitable majors and the lack of an integrated data management system. This study developed a decision support system based on the Profile Matching method to provide objective major recommendations by aligning students’ competencies with the criteria of each major. The system was developed using the Waterfall model to ensure a structured design process. Testing was carried out with White Box Testing to validate algorithm calculations, Black Box Testing to check system functionality, and User Acceptance Testing to assess user satisfaction. The results showed a very positive response from administrators (MOS 4.95 out of 5.00) and a positive response from students (MOS 4.00 out of 5.00). This research contributes a practical and flexible decision support solution specifically adapted to high schools implementing the Merdeka Curriculum. With this system, the major selection process is expected to become more efficient, integrated, and capable of providing recommendations that better align with students’ interests, talents, and academic potential.

Key words: Education, Decision Support System (DSS), Majoring, Profile Matching, Waterfall.

Multidimensional in Integrated Islamic Schools Based on Web Technology: Integration of Al-Qur’an Learning Modules and Daily Activities

Muhammad Yusuf Baihaqi Solihin Solihin — 2025-12-31

Currently, the manual system of report card assessors is no longer considered adequate for handling the workload. The process of assessing student grades on SDIT Anak Sholeh Mataram report cards which still uses conventional methods causes the assessment service process to be less than optimal. Therefore, to overcome these difficulties, a web-based e-report information system was designed using the PHP programming language and using the MySQL database to process report card filling data. This web-based e-report is not only used to fill in academic grades but there are other assessments that can be used, namely assessments in the context of tahfiz and tahsin and assessing students' daily activities. The method used in web development this web-based e-report information system is the rapid application development (RAD) method with several
stages of needs analysis, system design, coding and testing, program implementation, and maintenance. This method is used because it can shorten the time in developing an information system than traditional methods. Based on the questionnaire with 10 questions given, there were 78% who agreed with the creation of the application.

Front-End Development and Data Fetching for Visualization of Mataram City Scout Data and Activities

Arvelio Renathan Budisantoso — 2025-12-31

Mataram City's scouting administration information system currently relies on manual processing like Google Forms and Excel that encounter challenges at the Scout Unit level up to the District Council. These manual processing result in data mistakes, reporting delays, and challenges for the Regency Council to obtain real-time information. This project aims to create a website-based information system using the Scrum methodology in order to centrally digitalize members, scoutmasters, and scout activities data management. Using Next.js and TypeScript, the authors develop a front-end and data retrieval from API Endpoints. This application to enhance user productivity, facilitate cross-level reporting, and provide effective data visualization. These benefits are supported by the User Acceptance Testing results, which show an average feasibility score of 81.74% in the "good" category across functionality, reliability, usability, and efficiency aspects. This confirms that the digitalization of scout administration is both ready to be implemented and aligned with user expectations.

A Systematic Review Factors Influencing Patient Utilization of Telepharmacy Services

Lalu M. Rifqi Azami S — 2025-12-31

The advancement of telepharmacy has significantly transformed the delivery of pharmaceutical services, particularly in regions with limited healthcare access. This systematic review aims to identify key factors influencing patients' utilization of telepharmacy services. A comprehensive literature search was conducted across PubMed, ScienceDirect, and Google Scholar, following PRISMA guidelines. Inclusion and exclusion criteria were applied to select relevant studies published between 2015 and 2025. Thematic synthesis of five eligible studies revealed five key factors affecting telepharmacy adoption: perceived practicality, efficiency, education level and age, knowledge, and trust. These factors align with the Technology Acceptance Model (TAM), highlighting the roles of perceived usefulness, ease of use, and psychological readiness. Strengthening digital literacy, infrastructure, and trust-building interventions is essential to enhance telepharmacy utilization, especially in developing countries like Indonesia. The findings provide valuable insights for designing patient-centered telepharmacy policies and services.

Keywords: Telepharmacy, utilization factors, patient acceptance, Technology Acceptance Model, healthcare access.

Development of an Ar Site Visualization Application to Enhance the User Experience in Integrating Conceptual Achitectural Designs of Houses into the Real-World Environment

Febriyanto — 2025-12-31

The rapid development of immersive technology, particularly Augmented Reality (AR), has significantly supported the digitization of infrastructure in the Fourth Industrial Revolution. In the property industry, conventional visualization media such as two-dimensional (2D) floor plans are often difficult for non-technical users to interpret. This study develops an Android-based application, AR Site Visualization, that converts 2D floor plans into interactive three-dimensional (3D) models in real time. The development applied the Multimedia Development Life Cycle (MDLC) framework with six stages, employing Unity, Vuforia SDK, and Lean Touch for feature integration. The 3D models were created using SketchUp and Enscape to ensure realistic visualization. Evaluation through Black Box Testing and formative assessments demonstrated that all features, including marker scanning, 3D visualization, rotation, zoom, and translation, worked effectively, achieving a feasibility score of 71.5%. The results indicate that AR Site Visualization enhances spatial understanding and improves design communication users.

Comparative Analysis of Proposed CNN Performance with CNN and Naive Bayes from Kaggle in ChatGPT Tweet Sentiment Analysis

Alwi Pratama — 2025-12-31

The rapid growth of social media platforms such as Twitter has led to an increasing demand for efficient sentiment analysis methods. This study focuses on the performance comparison of the CNN-based sentiment
analysis model developed by the authors with two models sourced from Kaggle; CNN model and Naive Bayes model. In addition, ChatGPT is used as a reference in discourse exploration and sentiment analysis strategy development. ChatGPT is used to answer user questions, generate code, revise journals and the like. Performance evaluation is done in terms of inference time and accuracy. The findings reveal that the CNN model developed by the authors achieves superior accuracy compared to the CNN model from Kaggle, while the inference time developed by the authors shows a significant difference with a much higher number when
compared to the Naive Bayes model from Kaggle. This analysis highlights the trade-off between efficiency and accuracy in sentiment analysis tasks and provides insights for selecting the right model based on current trends in data analysis.

Multitask Aspect-Based Sentiment Analysis of Indonesian Tweets on Mandalika Circuit using CNN and IndoBERTweet Embeddings

Raissa Calista Salsabila — 2025-12-31

This study proposes a multitask Aspect-Based Sentiment Analysis (ABSA) model for Indonesian tweets related to the Mandalika Circuit, using IndoBERTweet embeddings and Convolutional Neural Networks (CNN). The model simultaneously predicts aspect categories and sentiment polarities. Two experimental setups were evaluated: one using raw tweets (Scenario 1) and another with preprocessed text (Scenario 2). The results show that Scenario 1 consistently outperforms Scenario 2, highlighting the ability of IndoBERTweet to handle informal tweet structures without requiring standard text cleaning. A paired t-test was conducted to evaluate statistical differences in performance between scenarios. While Scenario 1 showed higher average F1-scores, the p-value (0.7178) suggests no statistically significant improvement across all classes. Further analysis reveals that certain classes, primarily neutral and positive sentiments, tend to perform worse than negative sentiments. Data augmentation was shown to improve recall and help the model handle underrepresented classes, particularly for “Ekonomi-Negative” and “Fasilitas-Negative” labels. The study highlights the importance of preserving informal language structures and utilizing data augmentation to enhance ABSA performance on real-world tweet data.

Classification of Throat Disease Using CNNs: EfficientNetB0 and ResNet50

Aliyah Fajriyani — 2025-12-31

Throat diseases are one of the global health issues. Early diagnosis could be an effective solution to prevent more severe throat disease. Automatic diagnosis based on medical images is possible to obtain by using Convolutional Neural Networks (CNN). This study employs two pretrained models namely ResNet50 and EfficientNetB0. The dataset contained 79 throat images divided to seven classes (normal, chronic laryngitis, acute pharyngitis, chronic pharyngitis, acute tonsillitis, chronic tonsillitis, and acute tonsillopharyngitis). The study was conducted in several scenarios and implemented gradually. First scenario, seven classes were merged into four classes (normal, pharyngitis, tonsillitis, and acute tonsillopharyngitis). Second scenario, four classes were combined into three classes (normal, pharyngitis, and tonsillitis). Third scenario, three classes were grouped into two classes (normal and illness). The results indicated that both the ResNet50 and EfficientNetB0 architectures achieved the highest performance in the third scenario (two classes). Both models showed identical evaluation matrics with accuracy of 91,67%, precision of 90%, recall of 100%, and F1-score of 94,74%. Furthermore, this study suggests that a dataset with numerous classes and limited data can be addressed by merging classes, thereby increasing the data size within each class.

Key words: Classification, Throat Disease, CNN, ResNet50, EfficientNetB0.