Bibliometric analysis: Learning using generative AI

This research aims to 1) analyze the number of articles per year and their types, 2) analyze the five articles with the most citations, 3) analyze keywords in the data collected, and 4) analyze the relationship between authors. Bibliometric analysis was used to help researchers study bibliographic content and citation analysis of each article taken from Harzing's Publish or Perish (PoP) 8 database. The data was taken from Scopus January 2018-2024 entitled, “learning using generative AI” in English. The maximum number of articles accessed is 500 articles. The collected articles are stored in CSV and RIS format. Articles were filtered according to analysis needs using MS Excel and VOSViewer. Twenty-nine documents from the Scopus database have been used in this research. The results show that 1) in 2022, there was one manuscript; in 2023, there were 27 manuscripts; and in 2024, there was one manuscript published. There are ten manuscripts published in conference proceedings, 14 articles in journals, and five manuscripts in edited books; 2) Suh, Popenici, Crosthwaite, Kim, and Chan wrote the 5 most cited articles; 3) Future research topics include GPT, student agency, learning approaches, higher education, state-of-the-art, reinforcement learning, opportunities, and challenges. Potential future research (novelty) could use keywords that are not yet widely used, including pedagogy, simulations, Socratic tutors, teaching methods, neuro-inclusive learning, and metacognition; 4) a network of writers outside the group consisting of Han, Ariel; Leu, U; Leu, Eunso; Lim, Cheoil; Kim, Hyeoncheol; Lee, Jeongji; Kim, Jiwon is the cluster with the highest relationship between nodes.