CAPSULE NETWORK SEVERAL YEARS LATER: A BIBLIOMETRIC ANALYSIS

Bibliometric analysis aims to identify research trends, spot research centers, and find research gaps. We conducted a bibliometric analysis of the Capsule Network method from 2018 to 2023. The research flow began with data collection from Google Trends, Open Knowledge Maps and Scopus. Google Trends was used to get a rough idea of interest over time and the most popular countries. Open Knowledge Maps analysis helped identify intersections between papers related to Capsule Network. The primary data was obtained from Scopus, which was analyzed using ScienceDirect to determine the subject area and Publish or Perish to obtain metadata. The metadata was then analyzed using VOSviewer to identify keyword associations and understand annual trends, the number of articles, and most citations. We also searched for research gaps connected to the Connected Paper website to identify topics that could be developed in the future. The analysis showed a decline in interest in the Capsule Network method. We obtained 170 papers identified in Scopus with the highest publication in 2019 with 50 papers, then decreased by 86% in the following years. Most publications are in IEEE Access with the dominant subject area being Computer Science. The research gap that we can propose is related to quantum computation. We suggest further research on quantum computing which has excellent potential for the future development of the Capsule Network method.