From the Perspective of Computational Communication: A Study on the Dynamics of Online Public Opinion in the Comment Sections of Weibo Regarding the 315 Gala (2021-2025)

Abstract

This study takes the comment sections of blue-V media official accounts related to the "3·15" Gala on Weibo from 2021 to 2025 as research samples. By compre-hensively applying methods such as word cloud analysis, sentiment computa-tion, IP location correlation, and topic categorization, it systematically explores the evolution of netizens’ comment content, dynamic emotional characteristics, regional sentiment differences, and location-based topic preferences over the five-year period. Integrating theories from computational social science and communication studies, the research reveals a diachronic shift in consumer rights protection public opinion "from physical infringement to digital in-fringement." It also identifies regional patterns such as "relatively moderate sentiment in eastern regions, more pronounced negative sentiment in central and western regions, and divergent regional topic concerns." The findings pro-vide data support and practical pathways for the dynamic optimization of mar-ket supervision policies, including differentiated regulation, sentiment-oriented response, and precise reaction to regional issues.