作者: Jizhe Xia , Chaowei Yang , Zhipeng Gui , Kai Liu , Zhenlong Li
DOI: 10.1080/13658816.2014.894195
关键词:
摘要: A variety of Earth observation systems monitor the and provide petabytes geospatial data to decision-makers scientists on a daily basis. However, few studies utilize spatiotemporal patterns optimize management Big Data. This article reports new indexing mechanism with integrated support Observation (EO) metadata for global user access. Specifically, predefined multiple indices (PMIM) categorizes heterogeneous queries based patterns, are various categories. structure, Access Possibility R-tree (APR-tree), is proposed build an R-tree-based index using query patterns. The was compared classic R*-tree in number scenarios. experimental result shows that generally outperforms regular supports better operation Global Obs...