热带地理

热带地理 ›› 2018, Vol. 38 ›› Issue (6): 751-758.doi: 10.13284/j.cnki.rddl.003087

• 论文 • 上一篇    下一篇

城市计算视角下的空间粗糙关联规则方法研究

廖伟华a，聂 鑫b   

  1. （广西大学 a．数学与信息科学学院；b．公共管理学院，南宁 530004）
  • 出版日期:2018-11-30 发布日期:2018-11-30
  • 通讯作者: 聂鑫（1983—），男，教授，博士，研究方向为土地资源管理，（E-mail）toefl678@163.com。
  • 作者简介:廖伟华（1975—），男，湖南耒阳人，副教授，研究方向为城市计算，（E-mail）gisliaowh@163.com；
  • 基金资助:

    国家自然科学基金项目（71403063、71763001）；广西重点研发计划项目（桂科AB18126007）

Spatial Rough Association Rules Method from the Perspective of Urban Computing

LIAO Weihuaa, NIE Xinb   

  1. (a. College of Mathematics and Information Science; b. School of Public Administration, Guangxi University, Nanning 530004, China)
  • Online:2018-11-30 Published:2018-11-30

PDF (PC)

171

赞

摘要：

同位模式表示不同类型的实体在空间邻域内共同频繁出现的规律，是城市实体空间关联的主要表达形式，但不能挖掘出指定实体的空间关联，需要寻找新的计算方法。在城市计算的视角下，通过引入粗糙集研究城市空间关联问题发现：1）该方法能把复杂的地理空间关联问题转换成信息决策问题，在信息决策表中计算城市实体之间的空间关联等拓扑关系，计算过程和结果可以挖掘城市行业之间的空间集聚和关联问题。2）通过属性约简得到属性核可以把高维空间数据降维，找到影响空间关联的重要因子。3）该方法拓宽了城市计算的理论方法体系和粗糙集方法的行业应用。最后，通过Python爬取南宁市城市服务业数据，进行方法的验证，计算结果与成熟的Apriori算法结果，以及南宁市服务业空间关联实际情况基本一致，证明了粗糙空间关联方法的可行性和正确性。

关键词: 空间关联, 粗糙集, 城市计算, 服务业, 南宁市

Abstract:

With the rapid development of E-commerce and smart city in china, it produces the explosive growth of the urban data. These data is an important data source for urban calculation. This study introduced rough sets to compute urban service entities with spatial reference coordinates. This study has done following studies in this paper. 1) It got distance near table in a given distance value for every spatial urban service entities, then related spatial entities table with distance near table. Then it got every spatial near entities of urban service and spatial urban service transaction database. 2) It got information decision table from spatial transaction database using SQL technology. It used every urban service name as attribute name in decision table. It used any service as decision attribute, others as conditional attribute, then it can get spatial association rules for this decision service and others’ service. 3) It got attribute core and attribute reduction of spatial decision attribute using rough sets concept and method, then got spatial urban service association rules based on its reduction. The main contributions of this study are as follows: 1) By introducing rough sets, the complex geospatial association problem is transformed into information decision problem, and the spatial association and other topological relations between urban entities are calculated in the information decision table. The calculation process and results can mine the spatial aggregation and association problems between urban industries. 2) Attribute kernel can reduce the dimension of high-dimensional spatial data and find the important factors affecting spatial association. 3) Broaden the theoretical methods of urban computing and the application of rough set method. Through Nanning City service industry data from Python crawling to verify the method, the results of the calculation to the mature Apriori algorithm results, as well as the actual situation of Nanning City service industry spatial association is basically consistent, proving the feasibility and correctness of the rough spatial association method.

Key words: spatial association, rough set, urban computing, urban service, Nanning City