大数据构建的赋分评价模型与功能区识别研究  

Research on scoring evaluation model and functional regions identification constructed by big data

作　　者：贾斐雪 闫金凤[1] 王甜 JIA Feixue;YAN Jinfeng;WANG Tian(College of Geodesy and Geomatics,Shandong University of Science and Technology,Qingdao,Shandong 266590,China)

机构地区：[1]山东科技大学测绘与空间信息学院,山东青岛266590

出　　处：《测绘科学》2021年第8期172-178,共7页Science of Surveying and Mapping

基　　金：国家自然科学基金重大项目(41890854);山东省重大科技创新工程项目(2019JZZY020103)。

摘　　要：针对如何解决各类兴趣点(POI)的用地面积差异并精确地识别功能区等问题,该文构建了一种新的功能区识别模型。使用网上调查的各类POI的面积数据等,对各类POI的面积大小所处的区间进行赋值,整理出各类POI的面积评分表,并通过对各类POI进行核密度分析得到每个兴趣点的核密度值。通过构建面积评分与核密度值二级赋分评价模型,增加了POI数据的可用性,实现快速获取城市空间结构。经过精度检验,Kappa系数为0.76,功能区识别精度较高。研究结果可用于区位熵指数、土地利用混合度、平均最邻近距离与标准差椭圆等分析,有利于规划人员掌握城市空间结构与进行科学决策。A new functional area identification model is constructed to solve the problems of different land area differences of various point of interest(POI) and to identify functional areas accurately. Using the area data of various POI surveyed online, the area of various POI was assigned to the interval, the area score table of various POI types was sorted out, and the kernel density value of each POI was obtained by the kernel density analysis. By constructing the model of area scoring and kernel density value secondary assignment evaluation, the availability of POI data was increased to achieve rapid acquisition of urban spatial structure. After precision test, Kappa coefficient was 0.76,the functional area recognition accuracy was high. The research results can be used for analysis of location entropy index, land-use mixed degree, average nearest neighbor distance and standard deviational ellipse, etc.,which is helpful for planners to grasp the urban spatial structure and make scientific decisions.

关 键 词：青岛市 兴趣点 城市功能区 核密度分析 

分 类 号：P208[天文地球—地图制图学与地理信息工程]