<![CDATA[Reading literacy is a fundamental competency that serves as a benchmark for the quality of primary education in Indonesia. The 2022 National Assessment conducted by the Ministry of Education, Culture, Research, and Technology revealed that most primary school students remain in the moderate to low proficiency categories, indicating the urgent need for data-driven strategies to improve literacy outcomes. This study aims to develop a predictive model for primary school students’ reading literacy scores by employing three algorithms: Neural Network (NN), Support Vector Machine (SVM), and Generalized Linear Model (GLM). The analysis followed the CRISP-DM framework, utilizing 2022 National Assessment data that includes school condition variables, availability of facilities, and related literacy indicators. The evaluation results indicate that GLM achieved the best performance, with R² = 0.988, MSE = 0.000322, and MAE = 0.014151, outperforming NN and SVM. This result indicates that the relationships between variables tend to be linear after preprocessing, making GLM more effective under the applied data transformation strategy. The implemented GLM model accurately predicted literacy scores on new data, demonstrating potential for adaptive learning module design and targeted resource allocation. These findings provide practical contributions for schools and policymakers in formulating more effective strategies to enhance reading literacy among primary school students in Indonesia. It is also important to consider that some predictor variables may have inherent relationships with the target variable, which can influence predictive performance.]]>
Copyrights © 2026
