In recent years, online travel service platforms such as TripAdvisor have been actively used by tourists. These services include user-generated content, which is vast and difficult to interpret manually. Several previous studies used user-generated content (e.g., social networking services and TripAdvisor) for tourism analysis. Most of these studies did not perform a systematic text analysis. In this study, we propose a method of analyzing this content to understand the characteristics of sightseeing attractions. Specifically, we analyzed the reviews of foreign tourists who visited Japanese sightseeing attractions. The review data were collected from TripAdvisor. First, a correspondence analysis was conducted to understand the similarities between sightseeing attractions. Next, a co-occurrence network analysis was conducted to derive the theme clusters for understanding the characteristics of sightseeing attractions based on the words in the review. Finally, individual analyses were conducted based on the description of the derived themes at each sightseeing attraction. The results of the analyses demonstrate that the proposed method is effective for comprehending the characteristics of each sightseeing attraction. The proposed method is useful when using user-generated content for tourism analysis.
