Aim: To explore the attitudes of healthcare professionals and non-healthcare professionals on anorexic behavior on social media.
Background: The significant function of attitude in alleviating anorexic behavior has been widely recognized. However, traditional methods often fail to capture patients' hidden emotions due to stigma and fear of judgment. Social media provides a novel platform for anonymously examining these behaviors and emotions, offering insights into anorexic behaviors that can enhance intervention strategies.
Design: This study has a qualitative design based on machine learning.
Methods: Data was collected from Zhihu, Weibo, and Xiaohongshu social media platforms up to 1 September 2024. This study method consisted of five steps: data collection, data cleaning, validation of relevance, sentiment analysis, and content analysis using the K-means algorithm.
Results: This study comprised 1099 comments, comprising 277,793 words. Non-healthcare professionals had seven emotions (good, happy, surprise, anger, disgust, fear, and sad) for anorexic behavior, and negative emotions were predominant. Healthcare professionals had three emotions (happy, good, and sad), and negative emotions were predominant. Healthcare professionals have a role deficit in recognizing negative emotions.
Conclusion: The study emphasizes the need for healthcare professionals to improve the recognition of negative emotions expressed by non-healthcare professionals/patients and develop data-driven interventions that address psychological barriers, fostering holistic patient care and improving outcomes.
Keywords: Artificial intelligence; attitude; medicine; qualitative study; social media.
© The Author(s) 2025.