Development and Verification of APM Algorithm for Mapping Urban Leisure Industrial Clusters
The efficient development of the leisure industry is increasingly important for the upgrading of the urban consumer economy and the improvement of human environment quality,but existing research has not been able to accurately capture the basic spatial distribution pattern and structure of the urban leisure industry.Most of the existing spatial clustering algorithms are applied to urban land use analysis,and there are fewer algorithms for the characteristics of urban leisure industry data.At the same time,existing algorithms have problems such as strong subjectivity in methodology and the inability to accurately identify the internal situation of the city,so they cannot be well-adapted to the characteristics of the sea quantization and fragmentation of urban leisure industry data.Now that points of interest(POI)have become a very important data source in urban leisure industry research,spatial clustering algorithms for POI data needs to be developed to provide more effective support for subsequent in-depth research.To this end,this paper develops an APM(agglomeration pattern mining)model for urban leisure industry clustering based on the leisure point of interest data of Guangzhou city,using algorithms such as effective accessibility of amenities and peak value,and also verifies the scientific validity and application value of the APM model.The final APM model captures 3170 urban leisure industry clusters within a 500-meter walking life circle and confirms that it can accurately obtain the spatial distribution and structural characteristics of urban leisure industry clusters through a two-fold verification.Generally,the APM algorithm can better identify the spatial cold and hot spot distribution of the urban leisure industry;locally,the APM algorithm can more scientifically identify the boundary conditions of multiple leisure consumption business circles within the city.In multiple inner-city representative areas,the clustering boundaries formed by APM algorithm have a greater overlap with the actual road and building directions,water system boundaries,and regional scope,and the clustering clusters are more in line with the actual situation and have more clustering credibility and validity.In addition,the APM algorithm can capture the rich and diverse business structure of urban leisure industry clusters.In the case of Guangzhou,the APM model captures the urban leisure industry cluster structure composed of nearly 50 sub-types within 8 types of leisure amenities.Amongst them,the strongest agglomeration core is the four combined business types of catering,beauty salons,clothing,shoes and hats,and sports.This study is a methodological innovation for the study of the leisure industry agglomeration mechanism and has made innovative advancements in algorithm accuracy,practical application,and visualization efficiency.Compared with the Fishnet method,it can more scientifically and accurately identify the boundaries of multiple leisure consumption business districts within the city,achieving efficient capture of urban leisure industry clusters.Compared with the homotopic model,it can present a multi-category urban leisure business structure,surpassing the limitation that existing research can only capture two types of business groups.
unban tourism and leisureindustry clustering patternspatial data miningclustering algorithmpoint of interestGuangzhou city
NETL
NSTL
万方数据
