Developments in related built environment domains have demonstrated the wide-ranging potential of open datasets to unify community analytical efforts and cultivate a more rigorous and critical urban science 17, 18, 19, 20, 21. This is particularly important in urban applications where the rationale behind the model’s predictions needs to be transparent and explainable. Good feature representation not only helps to improve model performance but makes it easier for domain experts and decision-makers to understand and interpret the results of GeoAI models. Towards advancing analytical and methodological innovation in urban networks, uniform, contextually comprehensive, and open spatial network datasets can serve as an invaluable resource for the urban research community. Moreover, current graph-based learning methods continue to prejudice a technical interpretation of urban streets based largely on the structural properties of networks, despite emerging evidence that graph algorithms learn from both structural and attribute-based features 16. Specifically, constraints in data consistency and interoperability, model explainability, and the feature representation of varied built environment features within networks make this a complex task 13, 14, 15. While significant progress has been made, the task of generalising machine learning methods to urban networks remains a critical challenge. Machine learning and predictive GeoAI offer numerous untapped opportunities to extract valuable insights from urban networks and expand existing use cases 9, 10, 11, 12. Presently, network analytics is employed to optimise decision-making procedures across all urban scales, ranging from coordinating city-wide vehicle fleets to the planning and design of active mobility systems 7, 8. Urban networks offer a powerful and intuitive lens to view, understand, and model the complexity of cities 1, 2, 3, 4, 5, 6. Urbanity aids various GeoAI and city comparative analyses, underscoring the growing importance of urban network analytics research. Accompanying the dataset is an interactive, web-based dashboard we developed which facilitates data access to even non-technical stakeholders. The dataset’s strength lies in its thorough processing and validation at every stage, ensuring data quality and consistency through automated and manual checks. We extract streetscape semantic features from more than four million street view images using computer vision. Our workflow enhances OpenStreetMap networks with 40 + high-resolution indicators from open global sources such as street view imagery, building morphology, urban population, and points of interest, catering to a diverse range of applications across multiple fields. We introduce the Urbanity data repository to nurture this growing research field, offering a comprehensive, open spatial network resource spanning 50 major cities in 29 countries worldwide. Urban network analytics has become an essential tool for understanding and modeling the intricate complexity of cities.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |