Island tourism offers distinct attractions compared to land-based travel, providing visitors with unique natural environments and independent cultures not found on the mainland. Due to its ability to facilitate various leisure activities and provide an escape from daily routines, island tourism is expected to maintain steady demand. However, it is also more vulnerable to change in climate and weather changes, making it susceptible to significant restrictions and direct impacts. Despite the growing interest in island tourism, existing research has primarily focused on island classification rather than strategies for effectively utilizing island tourism resources. This study aims to develop a data-driven evaluation and activation strategy for island tourism resources in Jeollanam-do, South Korea. To achieve this, an island classification system is proposed to systematically categorize islands based on their tourism-related characteristics. The classification system is constructed using exploratory factor analysis and a clustering technique based on the Gaussian Mixture Model (GMM), allowing for the identification of key island attributes. Furthermore, an island recommendation algorithm is introduced based on the established classification system, providing strategic insights for th...