<![CDATA[Dropout in island regions such as Giligenting District is a crucial issue influenced by geographical and academic constraints. This study aims to predict the potential dropout risk among junior high school students using data mining techniques with the C4.5 algorithm. The dataset used consists of 358 student records covering demographic, academic, social, and economic attributes. The research stages include preprocessing, attribute weighting, and classification using RapidMiner with an 80:20 split data validation scheme. The testing results show that the model achieved an accuracy of 62.5 percent, precision of 68.42 percent, and recall of 76.47 percent. Based on attribute weight analysis, the most dominant factors influencing dropout risk are Average Grade and Distance from Home to School, followed by Attendance and Family Dependents. This study contributes as a foundation for an early warning system, enabling schools to carry out priority interventions for students with low academic indicators and long travel distances to school.]]>
Copyrights © 2026
