cover
Contact Name
muhammad siddik hasibuan
Contact Email
mhdsiddikhasibuan@gmail.com
Phone
-
Journal Mail Official
cosieaira@gmail.com
Editorial Address
Jl Pukat Banting IV NO 41 Medan Kecamatan Medan Tembung Kode Pos 20224
Location
Kota medan,
Sumatera utara
INDONESIA
Journal of Computer Science and Informatics Engineering
ISSN : -     EISSN : 28278356     DOI : -
Core Subject : Science,
Artificial Intelligence Machine Learning Natural Language Processing Computer Vision Text Speech Text Mining Data mining Cryptography Data visualization Expert System Deep Learning Fuzzy Logic IoT and smart environments Neural Networks Pattern Recognition Image Processing Optimization Digital Signal Processing Networking Technology Web intelligence
Articles 147 Documents
Implementation of Random Forest Algorithm for Web-Based Toddler Nutritional Status Classification Muhammad Raddhyan; Deanna Durbin Hutagalung
Journal of Computer Science and Informatics Engineering Vol 5 No 4 (2026)
Publisher : Ali Institute of Research and Publication

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.55537/cosie.v5i4.1840

Abstract

Stunting remains a persistent public health challenge in Indonesia, affecting 21.6% of children under five according to the 2022 Indonesian Nutrition Status Survey (SSGI), while nutritional status classification at Posyandu Makmur 1 is still performed manually using the WHO z-score table, a process prone to human error. This study integrates a Random Forest algorithm, optimized through 5-fold Grid Search Cross-Validation, to classify toddler nutritional status into four categories (severe malnutrition, undernutrition, normal nutrition, and overnutrition) based on anthropometric data, and implements the resulting model into a Flask-based web application developed using the Rapid Application Development (RAD) method. The model was trained on 263 records of toddlers aged 0-59 months (80:20 train-test split) and evaluated on 53 test samples, achieving 83.02% accuracy with n_estimators=200 and max_depth=15. However, additional evaluation using macro F1-score (59.90%) and balanced accuracy (61.64%) reveals a substantial performance drop for minority classes, particularly undernutrition, due to class imbalance in the dataset. Feature importance analysis identified height (35.5%) and age (30.1%) as the most influential predictors, consistent with the height-for-age Z-score formula used for data labeling. The resulting system passed both White Box and Black Box testing, confirming its functional readiness as an early screening tool at primary health services, and underscoring the need for class imbalance mitigation strategies in future machine learning-based nutritional classification research
Obesity Level Classification Using the Gaussian Naive Bayes and 10-Fold Cross-Validation Rivaldi Prima Nanda; Muhammad Siddik Hasibuan; Mhd. Ikhsan Rifki
Journal of Computer Science and Informatics Engineering Vol 5 No 4 (2026)
Publisher : Ali Institute of Research and Publication

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.55537/cosie.v5i4.1855

Abstract

Obesity classification is an important task in health-related data analysis for identifying obesity levels based on individual characteristics. This study aims to implement the Gaussian Naïve Bayes algorithm for obesity level classification and evaluate differences in performance estimates obtained using a 70:30 train-test split and 10-Fold Cross Validation. The dataset consisted of 1,000 observational records containing five predictor variables: age, height, weight, body mass index (BMI), and physical activity level. Obesity levels were categorized into four classes: Underweight, Normal, Overweight, and Obese. The model was implemented using Python and evaluated through two validation scenarios. Experimental results showed that the Gaussian Naïve Bayes model achieved an accuracy of 95.33% using the train-test split approach. Evaluation using 10-Fold Cross Validation produced an average accuracy of 96.30%, precision of 96.66%, recall of 96.45%, and F1-score of 96.48%. These findings indicate that different validation strategies can produce different performance estimates. On the dataset used in this study, Gaussian Naïve Bayes demonstrated high classification performance, while 10-Fold Cross Validation provided performance estimates across multiple data partitions, offering a broader assessment of model performance. This study provides empirical evidence regarding the use of different validation strategies for obesity level classification
Local Binary Pattern for Facial Gender Identification Using a Kivy-Based Application Wira Pranata Sebayang; Tomy Satria Alasi; Irwan Jani Tarigan
Journal of Computer Science and Informatics Engineering Vol 5 No 4 (2026)
Publisher : Ali Institute of Research and Publication

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.55537/cosie.v5i4.1858

Abstract

Automatic gender identification based on facial images still faces challenges due to spatial variations and the complexity of visual features, while deep learning approaches such as Convolutional Neural Networks require high computational resources that are unsuitable for resource-constrained devices. This study aims to design, implement, and evaluate the performance boundaries of the Local Binary Pattern method for facial image-based gender identification integrated into a Python application using the Kivy framework. The research methodology utilized facial images from the UTKFace dataset, undergoing preprocessing through grayscale conversion, micro-texture feature extraction via the Local Binary Pattern operator, and gender classification using a threshold-based decision rule, all of which were fully integrated into the Kivy user interface. The system was evaluated across three scenarios with varying numbers of test images (80, 160, and 320 images). Testing results revealed a peak accuracy of only 43%, with variations in lighting, facial pose, and expression identified as the primary factors limiting overall accuracy. It can be concluded that while the Local Binary Pattern method was successfully integrated technically within the Kivy framework, its simplified mean-value feature representation remains insufficient for reliable gender identification under uncontrolled visual conditions, highlighting the need for richer feature representations in future research.
Web-Based New Student Admission Information System for Applicant Prediction Using Linear Regression Nasrul Abdul Karim
Journal of Computer Science and Informatics Engineering Vol 5 No 4 (2026)
Publisher : Ali Institute of Research and Publication

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.55537/cosie.v5i4.1877

Abstract

At present, SMA Islam Hidayatul Ahkam still handles its New Student Admission (Penerimaan Peserta Didik Baru/PPDB) procedure by hand. Because of this, checking submitted documents takes a long time, data entry mistakes are common, and school staff struggle to estimate how many applicants to expect in the following intake for capacity planning purposes. This study sets out to design and build a web-based PPDB information system equipped with a Simple Linear Regression prediction module, taking SMA Islam Hidayatul Ahkam as the case study. The system was constructed following the Rapid Application Development (RAD) approach and implemented with native PHP, PDO, and a MySQL database. Four years of historical applicant records were processed using the least squares technique, yielding the regression equation Ŷ = 36 + 10.9X. From this equation, the system forecasts 91 applicants for the 2026/2027 academic year, with an operational range of 77 to 105 (a ±15% margin, not a formal statistical interval), a Mean Absolute Percentage Error (MAPE) of 15.14%, and an R² value of 47.65% (in-sample). Black box and white box testing confirmed that every system function operated correctly and that the automatically generated calculations matched the manual ones exactly, confirming software implementation consistency rather than the statistical validity of the forecast. These findings indicate that the resulting web-based PPDB system serves not merely as an administrative tool but also as a prototype predictive-analysis aid that still requires further validation, for planning and decision-making in new student admissions at SMA Islam Hidayatul Ahkam
Sentiment Analysis of the Free Nutritious Meals Program on the X Platform Using Naïve Bayes and SVM M. Jagad Kesuma; Ari Muzakir
Journal of Computer Science and Informatics Engineering Vol 5 No 4 (2026)
Publisher : Ali Institute of Research and Publication

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.55537/cosie.v5i4.1928

Abstract

 The Free Nutritious Meals (MBG) program is a government program aimed at supporting nutritional fulfillment and improving human resource quality. The high level of public interaction on X makes the platform a potential source of data for identifying public responses to the implementation of MBG. This study aims to analyze public sentiment toward the MBG program and compare the performance of the Naïve Bayes and Support Vector Machine (SVM) algorithms. The research data consisted of 5,048 Indonesian-language tweets collected through Browser Automation Scraping. The data were processed through cleaning, case folding, normalization, tokenizing, stopword removal, and stemming, followed by sentiment labeling into positive, neutral, and negative categories using a lexicon-based approach. The dataset was divided using an 80:20 ratio into 3,810 training data and 953 test data. SMOTE was applied to the training data to address class imbalance. The results showed that Naïve Bayes achieved an accuracy of 35.36% before SMOTE and increased to 70.20% after SMOTE. SVM with SMOTE achieved an accuracy of 85.73%, correctly classifying 817 of 953 test data, which was 15.53 percentage points higher than Naïve Bayes with SMOTE. These results indicate a difference in the performance of the two algorithms, with SVM achieving higher overall classification performance on the research dataset.
Design and Implementation of an Android-Based Student Stress Level Prediction System Using Naïve Bayes Nidia Rintriana; Uliyatunisa Uliyatunisa
Journal of Computer Science and Informatics Engineering Vol 5 No 4 (2026)
Publisher : Ali Institute of Research and Publication

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.55537/cosie.v5i4.1931

Abstract

Students frequently face academic pressure that impacts their learning performance and mental health. The process of determining student stress levels at Universitas Pamulang is still carried out conventionally and is not yet integrated, resulting in a time-consuming process and subjective conclusions. This research focuses on the design and implementation of an Android-based student stress level prediction system to solve these problems. The main method applied in the system development is the Naïve Bayes algorithm with the Laplace Smoothing technique to resolve zero probability values. The questionnaire consists of 15 questions and has been validated by a psychology expert. The functional reliability and internal structure of the software are validated using Black Box and White Box testing, whereas the predictive performance validity of the model is evaluated separately using a confusion matrix against 60 test samples (comprising 5 low, 32 moderate, and 23 high-risk samples). The evaluation results indicate that the system is able to classify stress levels with an overall accuracy of 85% (with class-wise F1-scores of 0.50 for low, 0.87 for moderate, and 0.88 for high categories), although the relatively small sample size limits the generalization of the findings to a broader population. Through this application, students can independently monitor their psychological condition as a preventive measure to manage stress and support their academic success in the university environment.
Implementation of JavaScript Object Notation in a Village Service Information System Abdul Rahman Ismail; Zainudin Sidik; Nur Oktavin Idris
Journal of Computer Science and Informatics Engineering Vol 5 No 4 (2026)
Publisher : Ali Institute of Research and Publication

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.55537/cosie.v5i4.1937

Abstract

 Village administrative services are a form of public service that requires data to be managed in a structured, fast, and accessible manner. Conventional service management practices may limit service submission, document management, and the delivery of information to the community. This study aims to implement JavaScript Object Notation (JSON) as a data exchange mechanism in a Village Service Information System that integrates a web-based application for village government and an Android application for the community. System development was carried out through three stages, namely system development, system testing, and system implementation. The system was developed using PHP and MySQL for the web-based application and Java for the Android application, with JSON used as the data exchange format between the database, API, web application, and mobile application. The implemented features include community registration and login, service submission, requirement document uploading, document validation by village administrators, approval and signature processes by the village head, downloading of service letters, and community complaint submission. Testing was conducted using Black Box Testing, White Box Testing, and JSON-based API request-response testing. The results showed that the main system functions, including login, service submission, service request data management, complaint management, and logout, operated according to the system design. The implementation of JSON also supported structured data exchange between the web and Android applications, enabling integrated service processes. Therefore, the implementation of JSON in the Village Service Information System can support data integration and the provision of web- and mobile-based village administrative services
Web-Based Quality Assurance Activity Monitoring Dashboard for LPM UIN Sumatera Utara Using RAD Zaki Mahbub; Diaz Alfaridzi; Arif Syafaruddin Gultom
Journal of Computer Science and Informatics Engineering Vol 5 No 4 (2026)
Publisher : Ali Institute of Research and Publication

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.55537/cosie.v5i4.1948

Abstract

 The Quality Assurance Institute (LPM) at UIN Sumatera Utara manages diverse quality assurance activities, but records remain distributed across printed documents and spreadsheets, hindering progress monitoring and report compilation. This study aimed to design and implement a web-based quality assurance activity monitoring dashboard using Rapid Application Development (RAD) that centralizes activity recording, monitoring, and reporting, and to verify its functional behavior against specified requirements. Development followed four stages: requirements planning, user design, construction, and cutover. Use case, activity, and class diagrams guided implementation. The backend used Node.js, Express, and Prisma with PostgreSQL, while the frontend used React and Vite. Functional evaluation employed black box testing of nine scenarios, one for each of nine functional requirements; non-functional requirements were specified as design requirements and were not empirically evaluated. The resulting system comprised nine web pages covering authentication, profiles, a summary dashboard, activities, a calendar, categories, documentation, documentation details, and user administration. All nine executed test cases (TC-01–TC-09) passed, yielding a functional test-case success rate of 100% (9/9) within the tested scope. The findings indicate that the implemented dashboard satisfied the specified functional requirements in the executed scenarios and brought quality assurance activity monitoring into a single web application. Its applicability to similar institutions remains to be validated. Effects on reporting time, usability, security, and operational performance were not measured and require further evaluation
UI/UX Design of a Web-Based UKS Information Service System Using Design Thinking Ali Ikhwan; Najwa Fadiyah Nasution; Salsa Fadila Nasution; Rohma Aulia
Journal of Computer Science and Informatics Engineering Vol 5 No 4 (2026)
Publisher : Ali Institute of Research and Publication

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.55537/cosie.v5i4.1951

Abstract

The School Health Service (UKS) is a form of health service provided within the school environment that plays a role in maintaining and improving students' health. Managing this service requires well-organized information delivery and a user-friendly interface. User Interface (UI) and User Experience (UX) aspects are crucial in system design, as they facilitate ease of use and effective user interaction. This study aims to design the UI/UX for a web-based School Health Service Information System using the Design Thinking method. The method comprises five stages: empathize, define, ideate, prototype, and test. The empathize stage identifies user needs and issues, while the define stage establishes the core problem. Subsequently, the ideate stage generates various solution alternatives, the prototype stage creates interface designs, and the test stage evaluates the designs using the Maze platform. The study results in a UI/UX design for the web-based School Health Service Information System. Testing was conducted using several task scenarios to measure user success in completing system activities. The findings demonstrate that the Design Thinking method can be effectively applied to create a UI/UX design centered on user needs. This design can serve as a reference for the development of the web-based School Health Service Information System
Integrated Executive Information System for Madrasah Education with Hierarchical Drill Down at the Ministry of Religious Affairs of North Labuhanbatu Regency M Arif Rahmat Pasaribu; Aninda Muliani Harahap
Journal of Computer Science and Informatics Engineering Vol 5 No 4 (2026)
Publisher : Ali Institute of Research and Publication

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.55537/cosie.v5i4.1964

Abstract

Madrasah education information at the Office of the Ministry of Religious Affairs of North Labuhanbatu Regency is spread across data groups with different characteristics, namely madrasah profiles, students, teachers and education personnel, infrastructure, and accreditation, which makes it difficult for executives to obtain a concise overview that can be traced down to sub-district and madrasah levels. This study aims to design and build a web-based Executive Information System (EIS) for Madrasah Education that integrates these five indicator groups into a single three-level Drill Down structure (regency–sub-district– madrasah). A qualitative approach using observation, interviews, and supporting data collection was applied to identify information needs, and development followed the Waterfall model using Laravel, MySQL, and the Unified Modeling Language. The reported contribution is an information structure design: a matrix of five indicator groups and three hierarchy levels that share one regional dimension, together with preprocessing steps that prevent double counting in aggregation. Black-box functional testing of eight scenarios shows that all scenarios passed (functional conformity of 8 out of 8, or 100%). These results demonstrate that the system functions as specified; they do not demonstrate improved executive decision quality, which was not measured and is left for future work together with usability evaluation.