Jurnal Ilmu Komputer dan Sistem Informasi (JIKSI)

Application of the Classification and Regression Tree (CART) Algorithm for Classifying Student Majors at MAN 1 OKU Timur

2025-06-27T17:00:09+00:00

The classification of new student majors represents a vital process to align students’ educational trajectories with their aptitudes and interests. MAN 1 OKU Timur, an Islamic senior high school under the Ministry of Religious Affairs (Kementerian Agama), categorizes students into three primary academic tracks: Natural Sciences, Social Sciences, and Religion. Traditionally, the classification process relied on a simple summation and ranking of academic scores, a method prone to subjective bias and misclassification. This study implements a data mining approach utilizing the Classification and Regression Tree (CART) algorithm to enhance the objectivity and precision of the major selection process. Using data from 379 students of the 2020/2021 academic year, the CART model achieved an accuracy rate of 88.65%, demonstrating high reliability in predicting appropriate majors based on students’ academic performance and stated preferences. The results indicate that the CART algorithm can serve as an effective decision-support tool for the classification of new students at MAN 1 OKU Timur.

Detection of Diabetic Retinopathy Using a Convolutional Neural Network (CNN) Algorithm

2025-06-27T17:00:18+00:00

Diabetic retinopathy is a disease that damages the blood vessels in the retina of the eye. If left untreated, this condition can lead to blindness. This study aims to detect and classify diabetic retinopathy using the Convolutional Neural Network (CNN) algorithm one of the deep learning methods applied in machine learning for image analysis and interpretation. The objective of this research is to enhance the accuracy of predicting and classifying the types of blindness experienced by diabetic patients based on retinal images. The system identifies four retinal conditions: normal retina, glaucoma, cataract, and diseased retina. The CNN model in this study was trained with an input image size of 2464×1632 using 90 training images and 10 testing images, a 3×3 filter, and 800 epochs. The model achieved 90% accuracy in classifying eye images during training and testing. The results demonstrate that CNN is highly effective in detecting diabetic retinopathy and differentiating between various retinal disorders.

Application of Data Mining Using Multiple Linear Regression to Project Population Figures at the Central Bureau of Statistics of Ogan Ilir Regency

2025-06-27T17:00:27+00:00

The Central Bureau of Statistics (Badan Pusat Statistik BPS) is a government agency mandated to provide comprehensive, accurate, and up-to-date population statistics to build a reliable, effective, and efficient national statistical system that supports national development planning. However, the BPS office of Ogan Ilir Regency faces challenges in its current population-data collection system, which remains incomplete. This study applies a multiple linear regression analysis method a multivariate technique used to estimate the relationships between dependent and independent variables of metric or non-metric type. The findings demonstrate that data-mining techniques are effective tools for addressing data-related problems by identifying useful and relevant knowledge within large databases. The proposed projection information system is designed to meet population-data requirements at the regency or municipal level for both present and future needs.

Analysis of End-User Satisfaction with the E-Learning System as a Learning Support Tool Using the EUCS Method at the Faculty of Tarbiyah and Teacher Training, UIN Raden Fatah Palembang

2025-06-27T17:00:36+00:00

The implementation of e-learning began in 2019 using a blended learning approach, and in 2020, the system was updated with new features, including virtual meeting capabilities. To ensure optimal service delivery, universities must maintain system performance by providing satisfaction guarantees for users, assuming that system stability has been achieved. However, there has been no systematic measurement of user satisfaction with the current e-learning system. This study involved 375 active students from the Faculty of Tarbiyah and Teacher Training to measure user satisfaction levels using the End-User Computing Satisfaction (EUCS) method, which includes five variables: content, accuracy, format, ease of use, and timeliness. The results revealed that the hypotheses for content, accuracy, ease of use, and timeliness were accepted, whereas the hypothesis for format was rejected.

Library Information System Using QR Code Technology at Patra Mandiri 2 Junior High School, Palembang

2025-06-27T17:00:45+00:00

The library of Patra Mandiri 2 Junior High School in Palembang still employs manual data collection methods, where records such as book inventories and transactions are maintained using notebooks and paper-based systems. This approach is time-consuming and prone to human error, which negatively affects library management and service efficiency. To overcome these limitations, a computerized Library Information System (LIS) was developed by integrating QR Code (Quick Response Code) technology as the central component for data access and management. The QR Code serves as a primary key for retrieving book information, while a webcam functions as the QR Code scanner. The research was conducted at Patra Mandiri 2 Junior High School, located in the Pertamina Complex, Flamboyan Street, Sungai Gerong, Plaju District, Banyuasin Regency, South Sumatra 30763. The study applied a descriptive research method and utilized the Rapid Application Development (RAD) model for system development. The objective of this study is to build a web-based library information system that integrates QR Code technology to facilitate borrowing, returning, and reporting processes. The resulting system enables efficient, accurate, and accessible library operations for students and staff.