| Abstract： |
“Small-Town Swot” has been a hot topic on the internet platforms in the past three years. Using the LDA topic model and SnowNLP sentiment analysis method, this paper conducted machine learning analysis on 38,369 Weibo posts containing the term “Small-Town Swot” to explore the generation and evolution path of public opinion. The results demonstrate that the core issues include “Individual Stress”, “Capital Gazing” and “Social Stratification”, with an increasing proportion of negative sentiment. “Individual Stress” is influenced by the uneven distribution of education and the impact of the COVID-19 pandemic, which results in a decrease in subjective well-being. The differences in individuals caused by “Capital Gazing” foster a sense of relative deprivation. The intertwining of subjective well-being and relative deprivation prompts discussions on social stratification. This paper clarifies the logic of network discourse generation of youth groups and effectively helps digital governance. In the future, the government should enhance the digitalization in rural towns to foster digital entrepreneurship and employment opportunities. It should also promote online educational resources to compensate for the inadequate access to education in rural areas and broaden career development paths. At the same time, the network monitoring system should be established to stabilize public opinion, enhance public trust and build social consensus. |
