Presentation Information
[E3-02]Uncovering the Relationship Between Tourist Attraction Features and Visitors Behavior Preferences: A Case Study of Deep Learning and Clustering Analysis with People Flow Data in Hakone
*Jianhao Shi1, Wanglin Yan1 (1. Keio Univ.)
Keywords:
Tourism,Spatiotemporal behavior,Deep learning,Clustering,K-means,LSTM
People flow and trajectory data provide key insights into policy adjustments and economic optimization, especially when combined with spatial information. However, research on this issue is yet limited because of the data availability. This study explores the relationship between the spatial features of tourist attractions and visitor spatiotemporal behavior preferences in Hakone Town.By employing a deep learning model with K-means clustering to high-density personnel flow data, the behavior patterns of tourists were identified first. These patterns were connected to the spatial characteristics of the town with LSTM (Long Short Term Memory) algorithm.The results show significant spatiotemporal differences in visitor behavior across various sightseeing spots of the town, with varying staying times and path complexities based on behavior preferences.