Business center is an important part of retail industry in urban spatial distribution, is an important carrier to identify the city center, and reasonable optimization of retail resource allocation can provide a reasonable reference for urban planning. The result is beneficial to the scientific layout of commercial space and the location of retailers. This paper takes the data of interest points (POI) in public Internet map as the research object, and looks for a method to identify urban commercial centers and retail formats; Taking Changsha as an example, this paper analyzes the spatial distribution of commercial activities and the concentration of retail formats. Research shows: According to the nuclear density estimation method, the commercial center is formed in the hierarchy of the central business center and the sub business center, and the results are in line with the objective facts. Through the nearest neighbor distance method, we find that the three types of commercial establishments are agglomeration spatial distribution, and the degree of agglomeration is less related to the quantity, from high to low respectively for convenience stores, shopping malls, commercial buildings and supermarkets. On the streets of the city as a unit, mainly distributed in Kaifu District and Furong district by hot region local Getis-Ord G* index method to identify the Changsha retail on the core pattern of industry development has been formed. Overall, Changsha retail business center and clustering analysis results can reflect the distribution characteristics of the commercial economy correlation distribution of the different retail agglomeration of regional space, the results are conducive to scientific planning of commercial space layout and retailer location.