Mining co-location patterns between network spatial phenomena

Abstract

The mining of co-location patterns is a popular issue in the field of spatial data mining. However, little attention has been paid to the co-location patterns between network spatial phenomena. This paper addresses this issue by extending an existing method to mining the co-location patterns between network spatial phenomena. The approach consists of two stages: (1) defining a co-location model on a network space based on skeleton partitioning of a road network to have co-occurrence relationships; (2) computing statistical diagnostics for these co-occurrence relationships. Our method was then applied to a case study regarding the mining of co-location patterns of manufacturing firms in Shenzhen City, China. These co-location patterns were also analyzed qualitatively according to the three mechanisms derived from agglomeration economies. Our method was compared with the existing method and the differences were verified by the network cross K-function.