结合网络评论数据与道路约束的城市兴趣域识别方法研究——以广州市越秀区为例  

Integrating Online Review Data and Road Constraints to Identify Area of Interest——A Case Study of Yuexiu District of Guangzhou City

作　　者：金兵兵[1] 宋颉 丁晶[1] 徐海滨 王楠[1] JIN Bingbing;SONG Jie;DING Jing;XU Haibin;WANG Nan(Guangzhou Urban Planning & Design Survey Research Institute, Guangzhou 510000, China;College of Geography and Environment, Shandong Normal University, Jinan 250358, China)

机构地区：[1]广州市城市规划勘测设计研究院,广东广州510000 [2]山东师范大学地理与环境学院,山东济南250358

出　　处：《地理信息世界》2019年第3期55-59,共5页Geomatics World

摘　　要：城市兴趣域的识别对城市研究具有重要的现实意义。当前识别大多是通过遥感影像以及实地调查,采用欧式距离的区域范围估计方法实现,没有考虑城市居民活动以及道路结构约束对兴趣域产生的影响。本研究提出一种结合网络评论数据和道路约束的城市兴趣域识别方法。首先,采用约束Delaunay算法,对空间结构复杂的道路进行简化;继而,设计基于Epanechnikov核函数的网络核密度算法,实现顾及道路约束的城市兴趣域空间范围的划分;最后,通过网络评论数据中的评论次数与评论文本,量化城市兴趣域的吸引程度,并依此判别其功能类型,从而实现兴趣域的识别。实验以广州市越秀区为例,成功识别出了宜安广场、北京路等城市兴趣域。该方法对城市空间结构的精细化识别、分析和规划具有重要的现实意义。The identification of area of interest(AOI) plays a significant role in urban studies. The most commonly used approaches to identify AOI only considered Euclidean distance-based region division are relied on remote sensing images or field survey, While the impact of human activities and road constraints on AOI have been ignored if not all. This paper proposes a novel approach to identify AOI by integrating online review data and road constraints.Firstly, the constrained Delaunay algorithm is used to simplify the complex road network. Then, the AOI regions has been extracted using network kernel density algorithm with a kernel function of Epanechnikov. Finally, the degree of attractiveness of AOI has been quantified by the counts of online review data, and the AOI categories have been defined using the textual information of online review data. Based on which, the AOI can be identified. Taking Yuexiu district of Guangzhou city as an example, the approach has effectively identified several AOI, such as Yi’an Square and Beijing Road. This study can shed lights on refined identification and analysis of urban spatial structure and urban planning.

关 键 词：城市兴趣域 网络评论数据 道路约束 

分 类 号：TP393.02[自动化与计算机技术—计算机应用技术]