소셜 빅데이터 기반 텍스트마이닝과 CONCOR 분석을 통한 광주광역시 관광 동향 분석

Abstract

This study aims to identify the structural patterns of tourist perceptions of Gwangju Metropolitan City by analyzing online social big data. Social media posts from major Korean platforms were collected over one year and processed using text mining procedures, including keyword extraction, frequency analysis, and network centrality analysis. A one-mode matrix was constructed, and CONCOR analysis was applied to classify structural equivalence among keywords. Sixty key terms were identified, with “restaurants,” “recommendation,” “hotel,” and “café” emerging as central nodes, confirming the dominance of gastronomy and urban leisure in Gwangju’s tourism image. The analysis further indicates that food-driven tourism acts as the primary anchor shaping visitors’ overall perception of the city. The CONCOR results revealed four upper-level clusters and seven detailed subgroups, including commercial-district-oriented, cultural-arts, seasonal-festival, and downtown-café clusters. These findings show that tourist perceptions consist of multidimensional structures integrating food culture, cultural assets, emotional consumption spaces, and mobility patterns. Practical implications include strengthening gastronomic branding, developing integrated culture–arts routes, enhancing regional leisure linkages, and improving stay and access infrastructure. This study provides a data-driven foundation for regional smart tourism curation and contributes methodologically through the integrated use of text mining and CONCOR techniques.