텍스트마이닝을 활용한 관광공기업의 ESG 경영전략 사례 분석: 지속가능경영보고서를 중심으로

This study empirically investigates the ESG(Environmental, Social, and Governance) strategy implementation structures of two leading public tourism enterprises in South Korea—Kangwon Land and Grand Korea Leisure(GKL)—through an analysis of their published sustainability reports. A three-step text mining methodology was employed, consisting of bigram-based keyword frequency analysis, Latent Dirichlet Allocation(LDA) topic modeling, and SentenceBERT (SBERT) embedding-based semantic similarity analysis. These methods were used to quantitatively assess the internal consistency between strategic objectives and their actual implementation. The findings indicate that Kangwon Land places greater emphasis on socially oriented strategies, particularly in the area of safety and human rights management. On the other hand, GKL focused more on environmental strategies such as expanding low-carbon management. By presenting an analytical framework for evaluating ESG management strategy in public tourism enterprises, this study contributes to the academic field. Moreover, it offers practical implications for improving the semantic alignment and expression consistency of ESG disclosures in the public tourism sector.