Guadix Martín, JoséLilic, MilicaRosales Martínez, Marina2024-12-112024-12-1120229781032226323https://hdl.handle.net/11441/165647Geographic data integration has been an ongoing problem when trying to approach multivariate analysis. Besides, workflows for new data analysis techniques and technologies in the Machine Learning and Big Data parallel computing domains remain a challenge for geographic data because of the intrinsic links between features due to topological relationships. Heterogeneous data structures, dissymmetry in data induced by the scale, and the geometric nature of geographic information, especially when using vector structures, hinder multivariate analysis. To overcome these difficulties, a new data structure, asymmetrical multiscale grids, and a companion software framework to build and manage them are proposed.application/pdf4engAttribution-NonCommercial-NoDerivatives 4.0 Internationalhttp://creativecommons.org/licenses/by-nc-nd/4.0/info:eu-repo/semantics/bookPartinfo:eu-repo/semantics/openAccess10.1201/9781003276609-5