The Application of the BERTopic Model in Natural Language Processing: In-Depth Text Topic Modeling

This study uses the BERTopic model, which is an advanced natural language processing (NLP) technology. The method combines BERT embedding for doing topic analysis. Our data source adopots social media text, to be more specific, we selected 881 answers from Zhihu, which is a famous Chinese social question-and-answer platform. The results extract and summarize seven main topics, revealing people's focus on different perspective of topics online. Among these, users pay great attention on health and mental health topics. In sum, this analysis technique provides a comprehensive perspective for understanding the public attitudes and concerns of users when analyzing specific events/issues. This study demonstrates the ability of advanced topic analysis techniques in analyzing social media texts and provides a powerful means and approach for understanding and exploring social phenomena in the digital age.