JURNAL PENELITIAN ILMU PENDIDIKAN
Vol. 18 No. 2 (2025): September-November

Bibliometric analysis of deep learning research trends as a pedagogical approach to science learning in elementary schools (2021–2025)

Handayani, Mega Setya (Unknown)
Warohmah, Mawaddah (Unknown)
Mahmudah, Umi (Unknown)



Article Info

Publish Date
26 Nov 2025

Abstract

The dynamics of education in the 21st century demand a pedagogical transformation towards holistic competency development. This study aims to analyze the research trend of Deep Learning as a pedagogical approach to social studies learning in elementary schools. Using bibliometric methods, this study mapped the intellectual structure of 574 scientific articles (2021–2025) sourced from Google Scholar through the Publish or Perish software and VOSviewer. The results of the network visualization analysis identified five main clusters that showed a shift in focus from the technical aspects of computing towards the integration of intelligent technologies in classroom practice. The publication trend jumped significantly after 2023, with the latest topics focusing on critical thinking, creativity, and artificial intelligence (AI) literacy. Density analysis reveals research gaps: the literature is dominated by science and STEM content, with little exploration of the social dimension (social studies) or local wisdom. This research concludes that Deep Learning has developed into a technological entity that supports the achievement of the 8 Dimensions of the Graduate Profile through the principles of mindful, meaningful, and joyful learning. The practical implications provide guidance for educators to integrate technology scaffolding with in-depth inquiry, shaping adaptive lifelong learners in the era of disruption.

Copyrights © 2025