系统仿真学报 ›› 2018, Vol. 30 ›› Issue (3): 840-845.doi: 10.16182/j.issn1004731x.joss.201803009

• 仿真建模理论与方法 • 上一篇    下一篇

Map Reduce框架下空间大数据的关联规则分析方法

张明智1, 李义1,2   

  1. 1.国防大学信息作战与指挥训练教研部,北京 100091;
    2.66194部队,北京 100012
  • 收稿日期:2016-04-01 出版日期:2018-03-08 发布日期:2019-01-02
  • 作者简介:张明智(1962-),男,陕西,博士后,教授,博导,研究方向为战争复杂系统建模与仿真、数据与规则;李义(1981-),男,河北,博士后,助工,研究方向为数据与规则、计算机战争模拟。
  • 基金资助:
    国家自然科学基金(61174156, 61273189),中国博士后科学基金特别资助(2017T100791)

Association Rules Analysis Method of Spatial Data Under MapReduce Framework

Zhang Mingzhi1, Li Yi1,2   

  1. 1.The Department of Information Operation & Command Training, National Defense University, Beijing 100091, China;
    2.The Forces of 66194, Beijing 100012, China
  • Received:2016-04-01 Online:2018-03-08 Published:2019-01-02

摘要: 空间数据具有空间性、时间性、多维性、大数据量、空间关系复杂等特点,如何在空间大数据里面寻找模式、规律和特征知识,用于战场态势感知和战场空间管理,这需要利用和开发一些非传统的数据筛选工具进行分析、挖掘。针对现有Apriori算法扫描数据库过于频繁的问题,结合Map Reduce的工作原理,对Apriori算法进行了改进,提出基于Map Reduce的空间大数据快速分析思路和技术框架,初步搭建了验证原型,对关键技术进行实验。实例验证结果表明,利用该技术路线和框架,可以提升海量空间数据的处理分析速度。

关键词: 空间数据, 大数据, 关联规则, 分析方法, 并行计算

Abstract: Spatial data has the characteristic of extensity, timeliness, multidimensional, large amount of data and complex relations. Some non-conventional data screening tool for analysis and mining is required to find out the patterns, rules and characteristics knowledge in the spatial big data for battlefield situation awareness and battle space management. In view that the existing Apriori algorithm scans the database too frequently, the Apriori algorithm is improved on the basis of working principle of Map Reduce .The fast analysis ideas and technologyframework of spatial data is proposed. An elementary validate prototype is built for the key technology experimentation.Experimental results show that, the technical route and framework can improve the speed of massive spatial data analysis and processing.

Key words: spatial data, big data, association rules, analysis method, parallel computing

中图分类号: