Research on territorial spatial planning based on data mining and geographic information visualization

• Published in: Nonlinear engineering Vol. 14; no. 1; pp. 103096 - 70
• Main Author: Wang, Yingqiao
• Format: Journal Article
• Language: English
• Published: Berlin De Gruyter 09.04.2025; Walter de Gruyter GmbH
• Subjects: Algorithms; Big Data; Classification; Cluster analysis; Clustering; clustering algorithm; Data mining; Decision support systems; decision tree; Decision trees; Economic development; Land use management; Land use planning; national space planning; Population distribution; Residential areas; Scientific visualization; Spatial data; time series; Urban development; Urban planning; Visualization
• ISSN: 2192-8029; 2192-8010; 2192-8029
• DOI: 10.1515/nleng-2025-0090

At present, big data and mining technology has been more and more applied to urban planning practice. Based on data mining technology and geographic information visualization, this article puts forward a study on the optimization of territorial spatial planning. This article selects the clustering algorithm and decision tree in data mining technology, combines them organically, and puts forward spatial data mining. Cluster analysis is used to conduct spatial cluster analysis of comprehensive spatial data to explore whether there is spatial correlation and large-scale contiguous data in the data; if so, it can realize the purpose of national territorial space division. According to the analysis of the relevant data of land use planning and the characteristics of the business process, the overall design of the land use planning decision support system was carried out. The preliminary effects of land use planning clustering were generated through the usage of the -central factor clustering algorithm, and the consequences with excessive classification accuracy have been chosen as coaching samples mixed with the baseline of time series. Finally, the nearest neighbor classification algorithm primarily based on dynamic time warping is used to classify and perceive the city's practical place again. With the useful resource of point of interest data, the remaining city characteristic identification result is obtained. Finally, the simulation analysis shows that the area under curve (AUC) value of seven types of land use is 0.9376 for cultivated land, 0.85442 for forested land, 0.81747 for grassland, 0.8708 for water area, and 0.86672 for rural residential area. The AUC value of other construction land is 0.80346, and the AUC value of urban construction land is 0.9376. In the urban expansion planning project, multi-source data such as land use status data, population distribution data, and economic development data are integrated, and the spatial agglomeration pattern of land use types is found by using the method proposed in this article, so as to provide a reference for rational planning of urban expansion direction.