Within the context of human–machine interaction, research on emotions has progressively emerged as a central focus. Over recent years, both academia and industry have endeavored to develop intelligent systems capable of perceiving, recognizing, and comprehending human emotions, thereby enabling machines to respond appropriately and naturally. However, among the current studies, researchers have mostly focused on a specific domain of human–machine emotional interaction, leaving few works that discuss the current state of emotional interaction research through a systematic and visual review of findings. Therefore, this article presented a four-stage research framework for conducting a systematic bibliometric review of emotion research in human–machine interaction. We reviewed 2846 publications from 2000 to 2025 and employed bibliometric methods to systematically analyze recent research trends and developmental trajectories in this field. Through in-depth examination of high-impact papers, author networks, core research institutions, and central research countries, this article outlined the academic landscape and collaborative ecosystem of emotion interaction research. Furthermore, we presented a framing and identified focus themes, including core concepts and theoretical foundations, algorithmic support for interaction technologies, implementation of multimodal interactions, intelligent machine and interface design, and application scenarios for interactions, which highlighted the current status of both main research topics and critical technological applications within the field of emotional interaction. This study also employed keyword burst analysis to sketch out future research directions, proposing to deepen research on multimodal interaction technologies and application scenarios centered on user experience, and further put forward suggestions for subsequent research from both human–side and machine-side perspectives. On this basis, we also introduced a conceptual model for future research on human–machine emotional interactions based on Predictive Mind. Overall, our research aims to provide scholars with a comprehensive perspective, fostering deeper exploration and innovation within the field of human–machine emotional interaction.
Guo et al. (Wed,) studied this question.