Literature Review 2020–2025: Deep Learning-Based Educational Game Media for Mathematics Learning in Elementary Schools

Abstract

Mathematics education in elementary schools plays an important role in developing students' critical thinking skills. However, the dominance of conventional methods often results in a superficial understanding of concepts, leading to poor learning outcomes. One innovative alternative is to integrate educational games with a deep learning approach to create a more effective and enjoyable learning environment. However, its implementation still faces obstacles, particularly the limited competence of teachers in managing learning. This study aims to conduct a systematic review of trends in the development of deep learning-based learning media focused on mathematics. The method used is a Systematic Literature Review (SLR) with the PRISMA framework of 37 articles published between 2020 and 2025, analyzed using VOSviewers. The results show an increasing trend in publications, mostly using experimental methods and R&D research. The educational games developed are generally digital-based and have been proven effective in improving students' critical thinking skills. The conclusion of this study confirms that the integration of educational games and deep learning is a relevant innovation with potential for application in mathematics learning in elementary schools.