Abstract
With the advent of the era of big data, geographic information system (GIS) is facing the challenge of massive, multi-source and heterogeneous data when processing and analyzing spatial data. This paper discusses how to use big data mining technology to optimize GIS spatial analysis methods to improve the efficiency and accuracy of spatial data processing. Firstly, the research emphasizes the importance of data preprocessing and cleaning, and ensures the quality of data through data integration, transformation and specification. Then, the paper introduces the process of spatial data mining in detail, including data import, feature extraction, and the selection and application of mining algorithms, such as cluster analysis, association rule mining and classification prediction. Through practical case study, this paper shows how to apply big data mining technology to identify the spatial pattern and temporal variation law of urban traffic flow, which provides a scientific basis for urban planning and management. Finally, this paper points out that although big data mining technology has made remarkable progress in GIS spatial analysis, it still needs to face challenges such as spatial heterogeneity, spatial scale effect, data integration and spatial data visualization. Future research needs to comprehensively use spatial statistics, machine learning, data mining and other methods, combined with domain knowledge for in-depth research and analysis. Through these methods, big data mining technology can be used more effectively to optimize GIS spatial analysis methods and provide more accurate spatial data analysis support for decision makers.