An exploration of research trends on metaverse: topic modeling with latent dirichlet allocation

Online platforms have supported users in collaborating and communicating with each other distantly. Adopting online platforms interconnected with the virtual world, especially the metaverse, has fostered interactive activities in diverse sectors. However, a deep understanding of how such platforms have been used and discussed in academia needs to be enriched. To tackle this issue, this study investigates the research trends of the metaverse. The search period was set to include all publication dates up to the analysis point on January 9th, 2022, so the research data of the papers published in the database until then was obtained. A total of 451 publications were collected and analyzed through the Latent Dirichlet Allocation algorithm of the topic modeling technique, exploring conspicuous topics, keywords, and pertinent publications over time. As a result, six topics were determined: Topic 1 (engaging in real and virtual worlds); Topic 2 (crypto marketplace); Topic 3 (teaching and learning in a virtual educational environment); Topic 4 (a figure of oneself traveling in a virtual world); Topic 5 (virtual marketplace reshaping retail); and Topic 6 (game-mediated activity in a virtual world community). The time series change in the number of publications on each topic was tracked, and an apparent increase was found since 2007. Further, keywords in each topic and relevant publications were obtained based on probability values and then elaborated. The findings illustrate what researchers have discussed regarding metaverse and suggest the direction for future study in the field.