The purpose of this paper is to understand tourism behavior in Yokohama by analyzing the tourist’s behavior on weekends and holidays in April and May of 2019 and 2022 using GPS data of visitors in Nishi and Naka wards of Yokohama City. The area was divided into 18 areas by tourist area, the transition data between areas were generated from the tourist movement, and these data were analyzed by the k-medoids method and Asymmetric Cluster analysis.
Analysis of the k-medoids method provided cluster structures centered on adjacent areas for both 2019 and 2022. In the asymmetric cluster analysis, clusters were constructed by relatively close areas in the anaylsis of the 2019 data, but in the analysis of the 2022 data, a characteristic cluster consisting of a wide range of areas from the Minato Mirai area to the Yamashita Park area centered on Shinko area appeared. It is shown that these results show differences in tourism behavior between the two sets of data. In addition, because of the analysis that considers asymmetry, the results allow for interpretation based on the directionality of tourisum behavior, which takes into account the inflows and outflows of the areas.
View full abstract