Abstract
With the increasing integration of artificial intelligence across various societal domains, research on human-AI collaboration has grown rapidly in recent years. This study adopts a thematic mapping approach based on the BERTopic modeling framework to examine the structural organization and evolutionary dynamics of the core human-AI collaboration literature from 2020 to 2025. The analysis identifies nine core themes, including creative collaboration, which are further grouped into four higher-level thematic clusters, such as "Cognitive and Knowledge Co-creation". Longitudinal results show that, although publication output has increased substantially over time, the overall distribution of thematic structures remains relatively stable. This stability suggests a pattern of proportional development across themes, providing a macro-level perspective for understanding the conceptual organization and evolution of the field.
Recommended Citation
Zhao, Yang; Li, Na; and Chen, Shuai, "Research on Human-AI Collaboration Based on BERTopic: Exploratory Thematic Mapping and Evolutionary Analysis (2020-2025)" (2026). ICEB 2026 Proceedings (Beijing, China). 71.
https://aisel.aisnet.org/iceb2026/71