Affective Computing in Education: A Bibliometric and Visualisation Analysis

The application of affective computing in education is gaining prominence as a key factor in enhancing learning outcomes. This study employs bibliometric analysis using CiteSpace to map 1,720 publications from the Web of Science Core Collection (2010–2024). The results reveal a three-stage evolution: theoretical foundation (2010–2015), technology deepening (2016–2019), and application explosion (2020–2024). While the field has transitioned from "emotion recognition" to "predictive intervention," a critical analysis through six educational lenses (including Vygotsky, Pekrun, and Noddings) exposes a persistent gap between technical capabilities and pedagogical needs. We argue that current systems prioritize algorithmic monitoring over genuine care, often neglecting student autonomy and social-emotional scaffolding. The study advocates shifting the research paradigm from "what technology can do" to "what education needs," providing guidance for responsible, learning-centered innovation.