Tourist Destination Segmentation in Jember Using Cluster Analysis for Data-Driven Tourism Development

Abstract

For strategic tourism planning in Jember, data-based segmentation must be applied in order to optimize destination management. The clustering of tourist attractions in Jember needs to be categorized in order to derive descriptive patterns and support the policies pertaining to tourism development. Experiments apply K-Means Clustering, DBSCAN, and Agglomerative Clustering methodologies to classify the tourist sites based upon name, rating, ticket prices, location, and reviews. Data Preprocessing involves encoding for categorical features, scaling for numerical features, and addressing missing values. Elbow Method, Silhouette Score, Davies-Bouldin Score, and Calinski-Harabasz Score are applied to ascertain the most fitting clustering solution. K-means clustering has identified three major clusters: high-priced premium destinations, low-priced mass spread destinations, and overwhelmingly crowded yet highly satisfactory destinations. DBSCAN shows two distinctive clusters and outlier clusters with unique destinations. Agglomerative Clustering gives clear separate clusters of high, middle, and low-cost attractions. Such findings are useful for tourism stakeholders to develop targeted marketing policies, improve visitor experiences, and use their resources in the most efficient way. By the performance indicators of each cluster, tourism managers can optimize service quality and competitiveness of the destination. This study assists in data-informed decision-making in regional tourism planning which goes hand in hand with sustainable tourism development.