位置:成果数据库 > 期刊 > 期刊详情页
网格环境下分布式空间离群挖掘体系的设计与应用
  • ISSN号:1560-8999
  • 期刊名称:《地球信息科学学报》
  • 时间:0
  • 分类:P208[天文地球—地图制图学与地理信息工程;天文地球—测绘科学与技术]
  • 作者机构:[1]福州大学福建省空间信息工程研究中心、空间数据挖掘与信息共享教育部重点实验室,福州350002, [2]福建省经济信息中心,福州350001
  • 相关基金:国家自然科学基金项目(30972299); 中-匈政府间科技合作项目(国科外字[2008]333号); 欧盟第七框架计划项目(FP7-2009-People-IRSES No.247608); 福建省重点科技项目(2010I0008)
中文摘要:

空间离群是指空间数据集中那些非空间属性值与邻域中其他空间对象明显不同的空间对象。空间数据一般按地理分布存储具有海量特性,传统的集中式处理模式不能满足海量数据处理的效率和空间数据本身的安全性等要求。因此,在研究小组开发的地理知识服务网格平台GeoKS-Grid的基础上,本文针对分布式空间离群挖掘,提出了一个基于网格的分布式体系框架,制定了网格环境下分布式空间离群挖掘的策略,实现了具体的分布式空间离群挖掘算法。另遵循分布式空间数据挖掘的一般过程和网格服务通用、可重用和可组合的原则,将算法按合理粒度进行分解,并封装成多个基本的原子服务,进而以网格工作流的方式进行服务发现与组合,完成包括局部离群挖掘和全局离群挖掘在内的分布式空间离群挖掘。最后,通过福建省生态地球化学调查土壤数据离群分析实例,验证了服务或系统的合理性和有效性。

英文摘要:

A spatial outlier is a spatial object whose non-spatial attribute values are significantly deviated from the other data's in the dataset.The identification of spatial outliers can lead to the discovery of some unexpected knowledge,and it has a number of practical applications.There are massive spatial data maintained over geographically distributed sites in WAN.It's necessary to analyse and process the data by using the high-performance distributed parallel processing system.Grid is one of the most effective approaches to meet this requirement.The geographical knowledge grid platform(GeoKS-Grid) established by our research group is the application of knowledge grid in geo-information science,which integrate technologies of grid computing,web service,WebGIS,data mining,information visualization,knowledge base of ontology and knowledge reasoning,online analytical processing,decision analysis,data warehouse and workflow,to form a geographical problem solving environment.In this paper,a grid based distributed framework and the corresponding strategy for distributed spatial data mining system are discussed,and a distributed algorithm for spatial outlier mining is designed and implemented.In general,the process of distributed spatial outlier mining can be seen to be a series of services including atomic services and composite services.Furthermore,according to the principle of web service reusage and compositionality,the distributed spatial outlier mining algorithm is decomposed into several grid atomic services.Distributed spatial outlier mining including local spatial outlier mining and global spatial outlier mining is realized by grid workflow approach to discovery and composition of knowledge atomic grid services provided by knowledge grid.Finally,demonstration application is carried out on the basis of soil geochemistry data inspected by the Ecological Geochemistry Survey of Fujian Coastal Economic Belt,the efficiency and the validity of the distributed spatial outlier mining service and system are verified

同期刊论文项目
同项目期刊论文
期刊信息
  • 《地球信息科学学报》
  • 中国科技核心期刊
  • 主管单位:中国科学院
  • 主办单位:中国科学院地理科学与资源研究所 中国地理学会
  • 主编:徐冠华
  • 地址:北京大屯路甲11号
  • 邮编:100101
  • 邮箱:sxfu@lreis.ac.cn
  • 电话:010-64888891
  • 国际标准刊号:ISSN:1560-8999
  • 国内统一刊号:ISSN:11-5809/P
  • 邮发代号:82-919
  • 获奖情况:
  • 国内外数据库收录:
  • 中国中国科技核心期刊,中国北大核心期刊(2008版),中国北大核心期刊(2011版),中国北大核心期刊(2014版)
  • 被引量:3181