ABSTRACT This study explores the distinct information needs of urban and rural populations by analyzing six months of Q&A posts on Naver's Knowledge‐iN in South Korea. Using Latent Dirichlet Allocation (LDA), KoBERT, and Non‐negative Matrix Factorization (NMF), we compared major themes within urban and rural posts. Our findings show that both groups share interests and concerns regarding dental healthcare, transportation, education, and food. Urban posts emphasized daily life services and mobile technology, reflecting interests in convenience and connectivity. In contrast, rural posts focused on regional welfare, local spots, and family or emotional concerns, suggesting possible service gaps and unique social dynamics. Topic distributions varied across the three topic modeling methods: LDA revealed broader categories, NMF highlighted more specific segments, and KoBERT captured context‐rich, nuanced themes. Overall, this comparative analysis underscores region‐specific information needs and demonstrates the complementary benefits of multiple topic modeling techniques for understanding social and digital inequalities.
Yang et al. (Wed,) studied this question.