Framing Community Happiness in Korean National and Local Newspapers: A Longitudinal Text-Mining Analysis, 2016–2025

This study examines how Korean national and local newspapers framed community happiness before, during, and after the COVID-19 crisis. Articles were retrieved directly from the BIGKinds news database by searching titles and full texts for the Korean equivalents of “community” AND “happiness” between January 1, 2016 and December 31, 2025. The portal returned 65,648 records; 63,214 were exported in six period-by-media Excel files, and 62,738 remained after removing 476 records flagged as duplicates or exceptions. The analytic corpus comprised 7,337 articles from 12 national dailies and 55,401 from up to 45 local dailies. Document frequency, TF-IDF, keyword co-occurrence networks, centrality and spectral clustering, and lexicon-based sentiment analysis were applied. National newspapers disproportionately emphasized expansion, children, the environment, facilities, the economy, and institutional support. Local newspapers more often linked happiness to gratitude, neighbors, care, older people, volunteering, and place-based relations. During 2020–2022, COVID-19 became the dominant network hub and degree centralization increased to .820 in national and .669 in local coverage. Mean sentiment fell sharply before recovering after 2023; local coverage remained more positive, although between-media effects were small. A common-window sensitivity analysis produced nearly identical keyword-gap rankings. The findings show that community well-being is publicly constructed through two persistent but complementary frames: an institutional-national frame and a relational-local frame. They also demonstrate crisis-induced agenda compression followed by post-crisis diversification.