A Text Mining Analysis of Media Coverage of Women’s Sports

Purpose: The purpose of this study is to explore how the meanings associated with women’s participation in sports have evolved within Korea by analyzing media coverage from 1995 to 2024. Method: A Python-based Jupyter Notebook environment was used for data preprocessing and analysis. News articles related to women’s sports were collected from the BigKinds database over a 30-year period, from January 1, 1995, to December 31, 2024. Word frequency analysis, term frequency–inverse document frequency (TF–IDF) analysis, network analysis, and CONCOR analysis were conducted using scikit-learn and other relevant Python packages. The final dataset consisted of 39,908 articles from 1995–2004, 54,176 articles from 2005–2014, and 81,270 articles from 2015–2024. Results: First, analysis of keyword frequencies revealed that terms such as match, athlete, Korea, victory, record, Olympics, golf, and gold medal were dominant throughout the study period. These findings suggest that media coverage of women’s sports has been largely shaped by themes of competition, performance outcomes, and medal achievements, reflecting a win-oriented perspective and a strong emphasis on elite sport. Second, CONCOR analysis identified a total of 18 clusters, with five clusters in Period 1 (1995–2004), six clusters in Period 2 (2005–2014), and seven clusters in Period 3 (2015–2024), indicating increasing thematic diversification over time. Conclusion: This study highlights the characteristics of social discourse surrounding women’s sports in Korean media and emphasizes the need to incorporate more multidimensional perspectives to support the sustainable development of women’s sports.