Comparison of K-means and K-medoids in Tourist Attraction Clustering based on Visitor Characteristics

The purpose of this study was to determine the clustering of tourist objects scattered in the Bangkalan area. Clustering has been done using data on the characteristics of tourist visitors. This approach is used to see how much influence the characteristics of visitors have on the attractiveness of tourist objects. We apply methods in data mining, especially algorithms for partitioning methods, namely K-Means and KMedoids. K-Means and K-Medoids were implemented in finding clusters of tourist attractions on the characteristics of gender, age, occupation, education, and marital status. The clustering results show that the K-Means and K-Medoids methods have their respective advantages on the characteristics of visitors. The best accuracy results are 90.48%, which is obtained on gender characteristics with the K-Medoids method