Deep learning to optimize literacy intervention with educational games in elementary schools

This study examines the impact of deep learning-supported game-based learning on literacy skills in primary school students in Indonesia. A four-month intervention involving 32 fifth-grade students used adaptive educational games to target phonemic awareness, vocabulary, and reading comprehension. Utilizing a quasi-experimental design with pre-test and post-test assessments, the study found an average literacy score increase of 31.84 points post-intervention. Students from lower socioeconomic backgrounds showed the greatest improvement, indicating the potential of adaptive, technology-assisted education to reduce learning disparities. The use of deep learning models to personalize feedback and adjust content to individual needs was key to enhancing student engagement. The findings suggest that integrating deep learning with game-based learning can significantly boost literacy outcomes, especially in under-resourced settings. Further research is recommended to evaluate these interventions across broader populations and extended timelines