Saying no to marriage: A topic modeling approach to Korean social big data on “bi‐hon”

Objective The purpose of this study was to identify latent topics in social big data related to bi‐hon , a Korean neologism referring to the civil status of unmarried individuals. Background Social big data can effectively capture complex and fast‐changing public discourse on the recent bi‐hon trend in South Korea and reveal the divergence in traditional marital formation and families. Method We analyzed 65,205 Korean online posts that mentioned bi‐hon and were posted on one of three online channels (news websites, Twitter, and online communities) from July 2016 to June 2019. Results Employing a topic modeling technique, we identified and labeled 20 latent topics that best represented the data collected from each online channel. Next, we categorized these topics into six overarching themes: (a) the changing landscape of marital formation in South Korea; (b) bi‐hon: a choice, declaration, and lifestyles; (c) gender inequality and feminism movements; (d) childcare burden in a country with the lowest fertility rate; (e) conflict with the Confucian tradition of Korean families; and (f) economic barriers to marriage. Conclusion Our study contributes to the growing scholarship on changes in marital formation and families around the globe by providing a comprehensive picture of bi‐hon and leveraging innovative social big data research. Implications This study calls for diverse policy support for the increasing single population and guides family researchers for future social big data studies.