A Bibliometric Analysis of Intelligent Voice Interaction Based on VOSviewer

VOSviewer was used as an analysis tool to sort out and analyze intelligent voice interaction. The time span was set as five years, and 12,558 results were searched in total. The database was screened by the types of papers in WOS core journal sets, and 2687 records were refined. Word frequency statistics and keyword clustering method based on author and country relationship were used to construct the atlas, and the clustering results were displayed intuitively by VOSviewer. The results show that the current studies of voice interaction mostly take behavioral experiments. Keywords of voice interaction include voice, recognition, perception, health, knowledge education, social participation, information retrieval, human-computer interface. Based on human-computer interaction research, artificial intelligence, behavior, gender keywords becomes new research trend in recent years. In the study of author distribution, Munteanu, Cosmin Clark, Leigh, Torre and Ilaria are the authors as the core, showing scattering pattern, and the correlation among authors is not very close. From the perspective of national branches, European and American regions led by the United States are the first to initiate research. In recent years, China, Spain, Australia and other countries have also emerged in the field of research.