Research report: Research on the evolutionary characteristics of hot topics in online public opinion of Chinese universities based on Weibo texts

University-related hot topics on Weibo between 2020 and 2024 clustered into seven categories and followed several distinct patterns, including cyclical fluctuations, event-driven bursts, changes influenced by educational values and polarized sentiment. Researchers collected and processed Weibo hot-search records related to universities across the five-year period. They used latent Dirichlet allocation, or LDA, to discover topics and the BERT language model to analyze sentiment associated with them.

Data visualization was then used to examine changes in the quantity of discussion and the sentiment surrounding each topic. This approach connected two measurements: the subjects attracting public attention and the emotional character of online responses to them. The analysis identified seven topic categories: female faculty and students; graduation and employment; campus management and security; education and teaching; illegal activities and violations; campus life; and examination and enrollment.

The quantity and sentiment of education-related public opinion showed cyclical fluctuations rather than one uninterrupted trajectory. Some topics were generated by unforeseen events, while shifts in prominent issues were influenced by educational values. These patterns are descriptive properties of the analyzed records; they do not establish the causes of particular changes or imply that every topic followed the same trajectory.

Sentiment polarization also emerged as a defining feature of the topic evolution. Online debates around some subjects produced sharply divided emotional reactions, showing that highly visible discussions could contain substantial disagreement. The study offers a methodological framework for tracking university-related public discourse, but its scope limits interpretation.

Because the research is observational, it cannot determine why individual events or educational values altered discussion. Nevertheless, its combined topic, sentiment and visualization analysis distinguishes recurring issue areas from event-driven bursts and maps emotional dynamics alongside changes in attention. The authors describe the approach as a reference for understanding and managing public opinion concerning Chinese universities, rather than as a forecast of future discourse.

Subject of Research: Evolution of university-related hot topics and sentiment in Weibo search data from 2020 to 2024

Article Title: Research on the evolutionary characteristics of hot topics in online public opinion of Chinese universities based on Weibo texts

Article References: He, S., Zhang, C., Wu, Z., & Yang, R. (2026). Research on the evolutionary characteristics of hot topics in online public opinion of Chinese universities based on Weibo texts. Humanities and Social Sciences Communications. https://doi.org/10.1057/s41599-026-09469-x

Image Credits: AI Generated

DOI: 10.1057/s41599-026-09469-x

Keywords: Weibo, Chinese universities, public opinion, topic evolution, sentiment analysis, BERT, LDA, social media research, online debates, higher education, Research, evolutionary