摘要
网络数据的广泛应用为城市商业空间研究提供了新方法和新手段。研究采集分行业商业POI数据、腾讯宜出行数据和商业用地数据,采用DBSCAN算法和核密度估计法,分析南京主城区的总体商业空间格局及分行业商业空间布局的特征。研究发现,南京主城区的总体商业空间呈单核集中块状结构,以新街口为中心随距离呈圈层式递减分布,分行业中仅商务和金融空间显现出多中心结构的发展趋势;南京主城区商业空间格局受到经济、政治、市场、规划、交通、企业等因素的共同作用和影响。建议各商业中心实现片区差异化、层次化、特色化发展,完善商业中心等级体系。
Network data provides new methods and new means for commercial space research. We explored commercial POI data, Tencent’s travel data and commercial land data. Using DBSCAN algorithm and Kernel Density Estimation, we conducted commercial spatial clustering analysis, and verified the results. Taking the central city of Nanjing as example, we explored the overall commercial space pattern and the characteristics of sub-commercial space layout. The study found that the overall commercial space is a single-center structure with Xinjiekou as the core, and that only business space and financial space shows the trend of multi-center structure. The spatial pattern of commercial space is influenced by economic, political, market, planning, transportation and enterprise factors. It is recommended that each commercial center realize the differentiation, layering and characteristic development and improve the hierarchy of the commercial center.
作者
翟青
秦萧
魏宗财
ZHAI Qing;QIN Xiao;WEI Zongcai(School of Geographic and Biologic Information,Nanjing University of Posts and Telecommunications,Nanjing 210023,China;School of Urban Planning and Design,Nanjing University,Nanjing 210093,China;School of Architecture,South China University of Technology,Guangzhou 510641,China;State Key Laboratoiy of Subtropical Building Science,South China University of Technology,Guangzhou 510641,China)
出处
《南京邮电大学学报(社会科学版)》
2019年第5期43-56,共14页
Journal of Nanjing University of Posts and Telecommunications(Social Science Edition)
基金
国家自然科学基金“城市居民网络在线活动对城市空间的影响机理及其效应研究”(41571146)
国家自然科学青年基金“购物行为破碎化作用下虚-实商业空间关联研究”(41601139),“保障房社区居民日常活动虚-实空间互动及其影响机理研究”(41801150)
江苏省自然科学青年基金“网络消费时代购物行为破碎化对虚-实商业空间的影响”(BK20160892)
南京邮电大学校引进人才项目“基于居民活动的城市实体空间和网络空间的互动影响机制研究”(NYY215016)
关键词
商业空间
商业规划
多源数据
DBSCAN算法
commercial space
business planning
multiple-source data
DBSCAN algorithm