Clustering of Tourist Zones and Facilities Using the DBSCAN Algorithm

This paper presents an analysis of the spatial distribution of tourist zones and facilities, and clusters tourist zones based on spatial density using the CRISP-DM framework to follow a structured process. The Data Understanding step was carried out using data on tourist attractions and facilities in Buriram and Nakhon Ratchasima provinces, obtained through web scraping. In the Data Preparation phase, the collected data was processed and structured to make it suitable for clustering analysis. In the Modeling phase, the DBSCAN algorithm was applied to identify spatially dense clusters of tourist zones. The Evaluation phase assessed the clustering performance using the Davies-Bouldin Index (DBI) and the Quality of Cluster (%). In the Deployment phase, the results were visualized through maps displaying the spatial distribution of tourist attractions and facilities, along with the clustering outcomes of the tourist zones. The research findings indicate that tourist zones in Buriram Province exhibit a high-density distribution, forming two distinct clusters, while tourist zones in Nakhon Ratchasima Province also show a high-density distribution with three distinct clusters. When comparing the clustering results with existing tourism routes in Buriram and Nakhon Ratchasima, it was found that the data from both sets are consistent, reinforcing the validity of the clustering method. This demonstrates that the DBSCAN algorithm can be effectively used as a tool for clustering tourist zones, supporting decision-making in tourism development and promotion.