Public perceptions and engagement with Traditional Chinese Medicine on Japanese social media (2010-2025): a text mining approach.

Traditional Chinese Medicine (TCM) is a holistic medical system whose global visibility has increased markedly, yet large-scale studies on public perceptions and engagement remain limited. Using Japanese Twitter data from 2010 to 2025, this study employs text mining techniques including BERTopic modeling, sentiment analysis, and co-occurrence network analysis to examine public discourse on TCM. Building on the Cognition-Affect-Behavior (CAB) framework, this study employs multilevel relational analyzes to examine the interplay between topics, sentiment, and behavioral engagement. The results show that public discussions mainly focus on clinical efficacy and daily wellness, traditional knowledge and scientific innovation, business and cultural promotion of TCM. Despite the dominance of neutral and positive sentiment, negative sentiment shows an increasing trend. Concerns are related to scientific rigor, ecological ethics, and adverse experiences. Japanese users are most engaged with tweets about TCM wellness, ingredients, and therapies. These behaviors reflect patterns of active learning, cultural identity, and experiential exploration. This study provides empirical support for the CAB pathway in social media environments and elucidates both the diffusion patterns of TCM-related discourse and the mechanisms underlying public responses on Japanese social media. It offers theoretical and empirical implications for cross-cultural health communication and the analysis of TCM discourse in digital contexts.