基于时空知识图谱的空间分析方法研究

A Spatial Analysis Method Based on Spatiotemporal Knowledge Graph

受数据量和数据种类影响,传统空间分析中跨图层运算问题是造成效率低下的关键因素。知识图谱作为一种新型的数据结构,"图"的思想能够有效应用于空间地理信息数据的存储。利用时空知识图谱框架,结合全球离散格网系统和Google S2算法将地理空间数据存储于地理格网中,以图谱检索的方式替代传统空间分析过程,降低了时间复杂度,可以有效提高空间分析检索效率。以便民生活圈中地理实体变化作为典型空间分析案例,比较了传统多图层空间分析和时空知识图谱检索方式的分析效率。结果表明,基于时空知识图谱的检索方法比传统多图层空间分析存在着一个量级的提升,这对于解决地理信息系统空间分析效率问题具有参考意义。

Limited by the amount and type of data,the cross layer operation in traditional spatial analysis results in low efficiency.As a new data structure,Knowledge graph could be effectively applied to the storage of spatial geographic data.Using the framework of spatiotemporal knowledge graph,combined with the discrete global grid systems and Google S2 algorithm,the geospatial data are stored in the geographic grid.The traditional spatial analysis process is replaced by graph retrieval,which reduces the time complexity and improves the efficiency of spatial analysis retrieval.Taking the changes of geographical entities in the convenient life circle as a typical spatial analysis case,this study compares the analysis efficiency of traditional multi-layer and spatiotemporal knowledge graph retrieval.The results show that compared with the traditional multi-layer spatial analysis,the method based on spatiotemporal knowledge graph retrieval has an order of magnitude improvement,which has reference significance for solving the efficiency problem of GIS spatial analysis.