Network-Constrained Spatial Point Pattern Analysis for Commercial Facilities

Based on social network reviews data and considering the structural features of the road network, we apply network kernel density estimation to study the spatial distribution pattern of commercial facilities.Meanwhile, the number and satisfaction of customers corresponding to the commercial facilities are used as a weighted index for network kernel density estimation, which reveals the spatial difference in consumer satisfaction of the commercial facilities.In some road sections, the spatial distribution of commercial facilities does not match the spatial distribution of the number and satisfaction of customers.In order to further analyze the spatial difference in consumer satisfaction of the commercial facilities, aimed at the structural features of the road network and the characteristics of the social network reviews data, this paper proposes network-constrained Getis-Ord G_i for quantitative analysis, The high-high road segment and low-low road segment are revealed, which reflects the spatial distribution patterns of the number and satisfaction of customers, and demonstrates the aggregated distribution of the satisfied or unsatisfied commercial facilities.These results will provide a quantitative reference for urban planning, commercial facilities layout, and location problem.