Papers
Retail Commercial Space Clustering Based On Post-Carbon Era Context: A Case Study Of Shanghai
Qinyu Cui Shenzhen University
Shuyu Zhang Tongji University
Yiting Huang Shenzhen University
Commercial aggregation and its influencing elements have obvious spatial heterogeneity in cities, and the study of its objective laws is an important part of urban planning and economic geography research, as well as an inevitable requirement for carrying out urban commercial network planning and scientific management. In the Post-carbon era, how to use multi-source urban data and machine learning to reasonably analyze and predict the site selection planning of commercial areas, and more accurate data-supported urban design can help to achieve resource-saving and maximize the rational allocation of businesses. From the literature review, we know that there is still a lack of research on this issue using new machine learning methods. In this paper, we take Shanghai, a city with a high level of economic and commercial development in China, as an example, and use the geographically weighted regression model (GWR) to analyze the spatial heterogeneity of the factors influencing the spatial clustering of retail businesses by combining multiple sources of data, and use machine learning algorithms to predict the location of future shopping areas in new planning areas. First, the 20 existing planned business districts in Shanghai are divided into three categories according to their size and geospatial location from the central urban area. Second, the commercial value (consumption level) of these three categories is taken as the dependent variable, and the geographical location (number of subway stations, number of bus stops, number of parking facilities, road network density), functional business (Points of Interests of schools, offices, hotels, restaurants, entertainment, etc.) of the business districts themselves and their surrounding areas, consumer population (number of people, daily visitation) are used as independent variables. Then, the geographically weighted regression model is established by using three types of indicators as independent variables: Further, a machine learning algorithm is used to predict the potential business districts in the future, so as to assess the commercial potential of the surrounding areas. The study shows that: (1) Shanghai still maintains the commercial spatial structure of the “municipal-regional-community level”, and there is significant spatial heterogeneity in the strength, positive and negative effects of different urban spatial elements on retail commercial spatial agglomeration, with a “center-periphery” structure in general. (2) Municipal commercial centers are influenced by factors such as people flow, subway construction, business offices, and attractions, etc. The factors that show a positive impact on regional commercial centers are, in order, people flow > residential > business offices > subway > schools, and the driving factors for community-level commercial centers are mainly public transportation and residential space. (3) There is spatial complementarity in the influence of subway and public transportation stations on commercial agglomeration. This paper helps to implement decent work and economic growth as well as responsible consumption and production. The data and analysis platform of this paper is supported by Metrodata Technology Company. (https://www.metrodata.cn/) Keywords: business district; business district hierarchy; agglomeration effect; spatial variability; geographically weighted regression model; megacities, machine learning, big data analysis; SDG8: Decent Work and Economic Growth SDG12: Responsible Consumption and Production Reference: [1]The correlation between HSR construction and economic development – Empirical study of Chinese cities[J]. Wangtu （Ato） Xu,Ying Huang. Transportation Research Part A . 2019 [2]Proximity to metro stations and commercial gentrification[J]. Jen-Jia Lin,Shu-Han Yang. Transport Policy. 2019 [3]Analysis on land ecological security change and affect factors using RS and GWR in the Danjiangkou Reservoir area, China[J]. Chaoxian Liu,Xueling Wu,Lei Wang. Applied Geography. 2019 [4]Land-use change, urbanization, and change in landscape pattern in a metropolitan area[J]. Hashem Dadashpoor,Parviz Azizi,Mahdis Moghadasi. Science of the Total Environment. 2019 [5]Interrelationships between retail clusters in different hierarchies, land value, and property development: A panel VAR approach[J]. Heeyeun Yoon. Land Use Policy. 2018 [6]Do build environments affect pedestrians’ choices of walking routes in retail districts? A study with GPS experiments in Hongdae retail district in Seoul, South Korea[J]. Yeankyoung Hahm, Heeyeun Yoon, Donggyu Jung, Hyunsook Kwon. Habitat International. 2017 [7]A Spatial Quantile Regression Hedonic Model of Agricultural Land Prices[J]. Philip Kostov. Spatial Economic Analysis. 2009 (1)
Keywords: Business District; Business District Hierarchy; Agglomeration Effect; Spatial Variability; Geographically Weighted Regression Model; Megacities, Machine Learning, Big Data Analysis; Sdg8: Decent Work And Economic Growth Sdg12: Responsible Consumption And