Identification of College Students’ Postgraduate Entrance Examination Situation Based on LDA Topic Modeling

Abstract

Understanding how university students make decisions about the Postgraduate Entrance Examination is important. It can help improve national education policies. This study uses data from social media platforms like Weibo, Zhihu, and Xiaohongshu. The study gathered comments related to the Examination through Python web scraping methods. After processing the data, the researcher applied the Latent Dirichlet Allocation (LDA) model to find 13 key topics. These topics cover policies and information, study planning, course reviews, and subsequent steps. The findings indicate that students consider both academic and non-academic aspects. They pay attention to threshold scores and academic requirements. Meanwhile, they also focus on psychological well-being and long-term career growth. Based on these insights, the paper suggests boosting the transparency of the Examination, creating psychological support systems for candidates, and advancing career planning education in universities. These actions are intended to gain a better understanding of the growth paths of today’s youth and to back sustainable social and economic development.