A Contribution to the Spatial Analysis of Territorial Systems Based on Graph-Structured Data

This work aims to present the methodological framework within which the T4Y project is developing as a contribution to the spatial analysis of complex systems related to tourism. The mapping of tourist ecosystems from a territorial perspective is a complex phenomenon due to the difficulty in understanding the interactions between tourists and the territory. Through the integration of AI-based techniques, the study aims to propose an approach based on graph-structured data to manage the complexity of geospatial information and improve planners' ability to abstract the possible relationships between elements of the territory. The approach utilizes graph-based data modeling while preserving spatial connections. It has been demonstrated how, through a bipartite graph approach, it is possible to decode the territorial components while taking into account the spatial connections between different pieces of information. Furthermore, the implementation of Graph Neural Networks (GNN) has been discussed as a way to facilitate a more integrated and context-aware analysis. This work aims to contribute to the ongoing debate on complex systems in spatial planning and provide valuable tools for policymakers and planners to experiment with scenarios for spatial development.