Urban area data are strategically important for public safety, urban management, and planning. Previous research has attempted to estimate the values of unsampled regular areas, while minimal attention has been paid to the values of irregular areas. To address this problem, this study proposes a hierarchical geospatial graph neural network model based on the spatial hierarchical constraints of areas. The model first characterizes spatial relationships between irregular areas at different spatial scales. Then, it aggregates information from neighboring areas with graph neural networks, and finally, it imputes missing values in fine-grained areas under hierarchical relationship constraints. To investigate the performance of the proposed model, we constructed a new dataset consisting of the urban statistical values of irregular areas in New York City. Experiments on the dataset show that the proposed model outperforms state-of-the-art baselines and exhibits robustness. The model is adaptable to numerous geographic applications, including traffic management, public safety, and public resource allocation.

Human interaction with indoor objects constantly changes indoor space and its navigability. Spatial subdivision models that delineate navigable space become a crucial prerequisite for indoor navigation. However, existing spatial subdivision models do not fully capture the dynamic changes of the indoor context and have limitations in identifying precise navigable space, thereby reducing the accuracy and efficiency of indoor navigation. This study proposes a novel interactive semantic model (ISM) that consists of an interactive semantic base (ISB) and empirical rules to accurately determine the navigability of the reshaped functional spaces (F-Spaces) of indoor objects. First, two-level F-Spaces (fine-grained resource F-Space and coarse-grained structure F-Space) are defined to express multi-granularity interactive semantics for delineating heterogeneous F-Spaces. Second, empirical rules are established through an extensible multi-dimensional semantics classification framework to determine each F-Space’s navigability. Lastly, a navigable F-Space generation scheme is designed by considering the adaptive navigability of the two-level F-Spaces. Simulated experiments show that the proposed model can generate precise and efficient dynamic navigable spaces. This study reduces the cognitive burden of human agents when facing indoor dynamic navigation, thereby improving the spatial experience of navigation.