摘要
在地学数据仓库的设计中,应当充分考虑地学数据自身的空间特点,采用区域维、时间维及属性维等多个维度,按不同主题所涉及的数据要求,进行相应的粒度与度量的定义,以建立不同特点的多维数据模型.以三峡库区地质灾害数据仓库的设计为例,对三峡库区时空范围各异的各类数据进行集成与融合,按5个主题对数据进行了归类.实验表明,该数据仓库可依不同灾害类型的、不同预警级别及不同分级对信息进行重组,能适应快速检索查询的需要;另一方面,利用库中的数据挖掘工具能在海量数据中挖掘地质灾害体的变化规律,从而使预测预报决策过程更加科学、准确、迅速.
Space-based design features eological data warehouse,unlike the traditional warehouse based on time.The key concept in the design of the multi-dimension model is to choose suitable multi-dimensions,such as spatial,temporal and attribute dimensions,and determine the granularity and measure,according the data requirements of the subject involved.A geohazards data warehouse of the Three Gorges Reservoir Area is built on 5 subjects to integrate the mass data in different space and time scales;the experiments show that the warehouse makes the geohazards prediction more accurate and rapid in two ways：Firstly,it enables fast search of the data since the data are re-organized by geohazards type,alarm levels and data layers;Secondly,it facilitates the discovery of the geohazards laws from data warehouse through the data mining tools.
出处
《地球科学（中国地质大学学报）》
EI
CAS
CSCD
北大核心
2011年第2期255-261,共7页
Earth Science-Journal of China University of Geosciences
基金
国士资源部"三峡库区地质灾害预警指挥系统(GHPACS)数据仓库及管理系统建设"(No.SXJC-3ZH1B1[2009]004)
关键词
数据仓库
地质灾害
三峡库区
数据挖掘
data warehouse
geohazards
Three Gorges Reservoir Area
data mining
