Evolution and regulation of online public opinion on Weibo: a corpus-based topic-sentiment aggregation analysis

Abstract This study investigates the evolution and regulation of online public opinion on social media Weibo, focusing on the “Enterprise N Incident” in the tea beverage industry. Using social media texts as a dataset, this study utilizes topic modeling and sentiment analysis to extract key topics and analyze their corresponding sentiment trends. Specifically, it explores how sentiment dynamics shift across different stages of public discourse, revealing the underlying patterns of opinion formation and transformation. Furthermore, it examines the interplay among emerging topics, sentiment fluctuations, and regulatory mechanisms, shedding light on how public sentiment aggregation is influenced by digital governance and platform policies. The findings contribute to a deeper understanding of online opinion evolution, offering implications for platform monitoring, crisis management strategies, and the development of more effective regulatory systems in digital communication environments.