Point process decomposition method for multi-scale spatial co-location pattern mining

doi:10.11947/j.AGCS.2022.20200548

Abstract

Abstract: Spatial co-location pattern mining aims to discover association rules formed by multiple types of geographic elements or events frequently adjacent to each other, which is the key for understanding the internal occurrence mechanism of complex geographic phenomena. Aiming at the shortcomings of existing spatial colocation pattern mining methods in the effective modeling of geographic data characteristics (such as the multi-scale characteristic), this paper proposes a multi-scale spatial co-location pattern mining method based on point process decomposition. Firstly, the spatial distribution of geographical elements with multiple types is modeled as a mixed spatial point process by constructing a random variable, and a non-parametric statistical index is introduced to discriminate the characteristic scale of the co-location patterns. On this basis, we define a conditional probability density distribution function to mine multi-scale spatial co-location patterns using points process decomposition. The experimental analysis results show that the proposed method can accurately depict the spatial distribution of spatial co-location patterns at different scales, and effectively reduce the subjectivity of artificially setting parameters.

Key words: multi-type geographical elements, spatial co-location patterns, point process decomposition, spatial multi-scales

CLC Number:

P227

DENG Min, CHEN Kaiqi, SHI Yan, CHEN Yuanfang, GUO Yiwen. Point process decomposition method for multi-scale spatial co-location pattern mining[J]. Acta Geodaetica et Cartographica Sinica, 2022, 51(2): 258-268.

References

[1] 刘文凯, 刘启亮, 蔡建南. 自然邻域支持下的空间同位模式挖掘方法[J]. 测绘学报, 2019, 48(1):95-105. DOI:10.11947/j.AGCS.2019.20170653. LIU Wenkai, LIU Qiliang, CAI Jiannan. Discovery of co-location patterns based on natural neighborhood[J]. Acta Geodaetica et Cartographica Sinica, 2019, 48(1):95-105. DOI:10.11947/j.AGCS.2019.20170653.
[2] BARUA S, SANDER J. Mining statistically significant co-location and segregation patterns[J]. IEEE Transactions on Knowledge and Data Engineering, 2014, 26(5):1185-1199.
[3] 李光强, 邓敏, 朱建军. 基于Voronoi图的空间关联规则挖掘方法研究[J]. 武汉大学学报(信息科学版), 2008, 33(12):1242-1245. LI Guangqiang, DENG Min, ZHU Jianjun. Spatial association rules mining methods based on Voronoi diagram[J]. Geomatics and Information Science of Wuhan University, 2008, 33(12):1242-1245.
[4] YOO J S, SHEKHAR S. A joinless approach for mining spatial colocation patterns[J]. IEEE Transactions on Knowledge and Data Engineering, 2006, 18(10):1323-1337.
[5] HUANG Yan, SHEKHAR S, XIONG Hui. Discovering colocation patterns from spatial data sets:a general approach[J]. IEEE Transactions on Knowledge and Data Engineering, 2004, 16(12):1472-1485.
[6] KOPERSKI K, HAN J. Discovery of spatial association rules in geographic information databases[C]//Proceeding of the 4th International Symposium on Spatial Databases. Portland:Springer, 1995:47-66.
[7] SHEKHAR S, HUANG Y. Discovering spatial co-location patterns:a summary of results[C]//Proceeding of the 7th International Symposium on Spatial and Temporal Databases. Redondo Beach, CA:Springer, 2001:236-256.
[8] Nilsson I M, Smirnov O A. Measuring the effect of transportation infrastructure on retail firm co-location patterns[J]. Journal of Transport Geography, 2016, 51:110-118.
[9] YOO J S, SHEKHAR S, SMITH J, et al. A partial join approach for mining co-location patterns[C]//Proceedings of the 12th Annual ACM International Workshop on Geographic Information Systems. New York, NY:ACM, 2004:241-249.
[10] XIAO Xiangye, XIE Xing, LUO Qiong, et al. Density based co-location pattern discovery[C]//Proceedings of the 16th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems. Irvine, CA:ACM, 2008:1-10.
[11] DENG Min, CAI Jiannan, LIU Qiliang, et al. Multi-level method for discovery of regional co-location patterns[J]. International Journal of Geographical Information Science, 2017, 31(9):1846-1870.
[12] CAI Jiannan, DENG Min, LIU Qiliang, et al. Nonparametric significance test for discovery of network-constrained spatial co-location patterns[J]. Geographical Analysis, 2019, 51(1):3-22.
[13] CELIK M, KANG J M, SHEKHAR S. Zonal co-location pattern discovery with dynamic parameters[C]//Proceedings of the 7th IEEE International Conference on Data Mining (ICDM 2007). Omaha, NE:IEEE, 2007:433-438.
[14] DING Wei, EICK C F, YUAN Xiaojing, et al. A framework for regional association rule mining and scoping in spatial datasets[J]. Geoinformatica, 2011, 15(1):1-28.
[15] QIAN Feng, CHIEW K, HE Qinming, et al. Mining regional co-location patterns with kNNG[J]. Journal of Intelligent Information Systems, 2014, 42(3):485-505.
[16] 蔡建南, 刘启亮, 徐枫, 等. 多层次空间同位模式自适应挖掘方法[J]. 测绘学报, 2016, 45(4):475-485. DOI:10.11947/j.AGCS.2016.20150337. CAI Jiannan, LIU Qiliang, XU Feng, et al. An adaptive method for mining hierarchical spatial co-location patterns[J]. Acta Geodaetica et Cartographica Sinica, 2016, 45(4):475-485. DOI:10.11947/j.AGCS.2016.20150337.
[17] 裴韬, 李婷, 周成虎. 时空点过程:一种新的地学数据模型、分析方法和观察视角[J]. 地球信息科学学报, 2013, 15(6):793-800. PEI Tao, LI Ting, ZHOU Chenghu. Spatiotemporal point process:a new data model, analysis methodology and viewpoint for geoscientific problem[J]. Journal of Geo-Information Science, 2013, 15(6):793-800.
[18] PEI T. A non-parameter index for differentiating between heterogeneity and randomness[J]. Mathematical Geosciences, 2011, 43:345-362.
[19] PEI Tao, ZHU Axing, ZHOU Chenghu, et al. A new approach to the nearest-neighbour method to discover cluster features in overlaid spatial point processes[J]. International Journal of Geographical Information Science, 2006, 20(2):153-168.
[20] KARLIS D, XEKALAKI E. Choosing initial values for the EM algorithm for finite mixtures[J]. Computational Statistics & Data Analysis, 2003, 41(3-4):577-590.
[21] MOHAN P, SHEKHAR S, SHINE J A, et al. A neighborhood graph based approach to regional co-location pattern discovery:a summary of results[C]//Proceedings of the 19th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems. Illinois, Chicago:ACM, 2011:122-132.
[22] YOO J S, BOW M. Mining spatial colocation patterns:a different framework[J]. Data Mining and Knowledge Discovery, 2012, 24(1):159-194.
[23] BYERS S, RAFTERY A E. Nearest-neighbor clutter removal for estimating features in spatial point processes[J]. Journal of the American Statistical Association, 1998, 93(442):577-584.
[24] HUANG Yan, ZHANG Pusheng. On the relationships between clustering and spatial co-location pattern mining[C]//Proceedings of the 18th IEEE International Conference on Tools with Artificial Intelligence (ICTAI'06). Arlington, VA:IEEE, 2006:513-522.
[25] ANDRESEN M A. The ambient population and crime analysis[J]. The Professional Geographer, 2011, 63(2):193-212.
[26] TATEM A J. WorldPop, open data for spatial demography[J]. Scientific Data, 2017, 4(1):170004.
[27] DUAN Kaibo, KEERTHI S S, CHU Wei, et al. Multi-category classification by soft-max combination of binary classifiers[C]//Proceeding of the 4th International Workshop on Multiple Classifier Systems. Guildford:Springer, 2003:125-134.