The Spatial Analysis and Site Selections of Retail Shops in Guiyang, China

摘要

Based on retails marketing data with geographic characteristics, this paper proposed a new combination of geography methods to analysis the spatial correlation between retail shops. And a spatial accessibility model was established to evaluate the commercial potential of each retail shop with social media data, which can indirectly grasp the consumer behavior characteristics in each region. Based on BP network, a hybrid spatial correlation evaluation model was established combined with social economy factors to find out the regions with highest market potential and some similar laws between them. We conducted the experiment in Guiyang, China, and considered 18 social economy factors to make the site selection more convinced. Through the experiment, 43 locations were found out with most business value and recommended to the new retail shops. The accuracy of the PCA-BP model was then proved to be satisfactory through the comparison with other regression methods. The model we established can guide retail chains to enhance business location planning and formulate regional development policy.