Planning and layout of tourism and leisure facilities based on POI big data and machine learning

Abstract

The spatial arrangement of tourism cities and the strategic placement of tourism and leisure facilities are pivotal to the development of smart tourism cities. The integration of Point of Interest (POI) data, enriched with location-specific insights, holds significant potential for urban planning and the optimization of spatial layouts. This study employs machine learning methodologies to evaluate the suitability of Beijing’s main urban area for the introduction of new tourism and leisure facilities. Drawing on POI and demographic data, and considering the distribution patterns of existing tourism and leisure facilities, this research applies machine learning to quantitatively simulate the optimal siting of such amenities. Key findings include: Firstly, compared with the existing tourism and leisure facilities, the fitting degree tested by the machine learning algorithm is 83.5%, suggests that the proposed method is highly feasible. Secondly, the decision-making model, trained with the CART algorithm, reveals that accommodation availability, shopping choices, and transportation infrastructure significantly influence the siting of tourism and leisure facilities in Beijing’s urban core. Thirdly, the model training indicates that facilities at various levels in Beijing exhibit a centralized layout, aligned with the city’s central axis, with a higher concentration in the urban center than in peripheral regions. The predictive analysis suggests that new tourism and leisure facilities are likely to be concentrated in densely populated areas. Lastly, some areas currently devoid of tourism and leisure facilities are identified as prospective sites for development. It is recommended that these areas be prioritized for the strategic placement. By leveraging machine learning algorithms for facility siting, this study aims to enhance the overall urban layout while mitigating the inherent subjectivity in planning and location decisions, offering valuable insights for the site selection of diverse facilities.