基于签到数据的城市热点功能区识别研究

Identification of Urban Interest Function Region by Using Social Media Check-in Data

城市功能是城市整体活动特点和类型的体现,识别城市内部功能区的空间分布结构,有利于城市结构优化,城市资源优化配置和城市发展规划等。以深圳市为例,对抓取的新浪微博位置签到数据进行了有效的数据筛选清洗,结合《城市用地分类与规划建设用地标准》和居民出行目的,对兴趣点（points of interest,POI）数据重分类,并针对分类完成的POI数据进行点要素空间多密度聚类,结合用户签到频率、POI数量比例以及土地利用混合程度构建了POI数据模型,以综合POI空间信息和语义信息,实现了城市空间自组织形态的功能区域主体功能识别。识别结果与深圳市城市布局结构规划图以及高德地图对比发现,该方法的聚类结果大体吻合深圳市空间布局的基本骨架,同时也能基本识别各大功能分区的主体功能和具有明显特征的功能区。

Taking Shenzhen city as an example,this paper combines the classification of city land use and standard of land use planing and construction land and the travelling purpose of residents to reclassify the points of interest（POI）,after cleaning the check-in data obtained from Sina Weibo.We apply a multi-density spatial clustering analysis method to those preprocessed location-based check-in data.Then,POI data model is constructed for the recognition of the main function of certain cluster,making ageneral consideration of the check-in frequency,the quantity proportion for a certain cate-gory of POI,and the diversity of POI category in a certain cluster.The validation is implemented by making contrast to the general and district planning in Shenzhen and the geoinformation of auto navi map.Results show that the clustering results of the proposed method are in general agreement with the basic framework of the spatial distribution of Shenzhen city,and this method can identify the main functions and the functional areas with obvious characteristics.