Sentiment and Topic Analysis of Autism Spectrum Disorder Discussions on Chinese Social Media: Evidence From Bilibili...

Social media has become an important platform for Autism Spectrum Disorder (ASD) communication. This study aimed to analyze online posts and comments related to ASD on Chinese social media, focusing on the emotional tendencies and topics about ASD.

This study used Python to retrieve ASD posts and comments, including 1,563 posts and 17,908 comments on Chinese social media. We employed content analysis to categorize posts and utilized the BERTopic model alongside Kernel Density Estimation (KDE) to examine sentiment distribution and thematic clusters.

(1) Content analysis revealed that in addition to Educational Information (44.5%), Personal Experience (25.5%), and Daily Life (23.1%), a fourth distinct category emerged in Chinese social media posts concerning ASD: Media Representation. This category captures how film, television series and novels shape public perception about ASD; (2) Sentiment analysis uncovered a stark "emotional polarization." While post content remained relatively balanced (52.8% positive), comment sections exhibited an overwhelming negative bias (66.9% negative), with sentiment scores heavily skewed toward the extreme low range (0-0.2); (3) BERTopic modeling identified four positive topics centered on family narratives, emotional support, intervention strategies and future survival planning. In contrast, the ten most prevalent negative topics emerged: half focus on conceptual controversies surrounding ASD, while the remainder address caregiving burdens, media criticism, and profound ethical conflicts regarding the life value of autistic individuals.

Social media in China plays an important role in public education regarding the conceptual understanding of ASD. However, these platforms provide limited emotional and informational support for affected families, while also perpetuating negative narratives and stereotypes that contribute to the stigmatization of ASD.