...
首页> 外文期刊>IEEE Transactions on Knowledge and Data Engineering >I/O-Efficient Bundled Range Aggregation
【24h】

I/O-Efficient Bundled Range Aggregation

机译:具有I / O效率的捆绑范围聚合

获取原文
获取原文并翻译 | 示例
           

摘要

This paper studies bundled range aggregation, which is conceptually equivalent to running a range aggregate query separately on multiple datasets, returning the query result on each dataset. In particular, the queried datasets can be arbitrarily chosen from a large number (hundreds or even thousands) of candidate datasets. The challenge is to minimize the query cost no matter how many and which datasets are selected. We propose a fully-dynamic data structure called aggregate bundled B-tree (aBB-tree) to settle bundled range aggregation. Specifically, the aBB-tree requires linear space, answers any query in (O(log _B N)) I/Os, and can be updated in (O(log _B N)) I/Os (where (N) is the total size of all the candidate datasets, and (B) the disk page size), under the circumstances where the number of datasets is (O(B)) . The practical efficiency of our technique is demonstrated with extensive experiments.
机译:本文研究捆绑范围聚合,从概念上讲等效于对多个数据集分别运行范围聚合查询,并在每个数据集上返回查询结果。特别是,可以从大量(数百甚至数千个)候选数据集中任意选择查询的数据集。挑战在于无论选择多少个数据集和最小化的数据集,都将查询成本降至最低。我们提出一种称为聚集捆绑B树(aBB-tree)的全动态数据结构,以解决捆绑范围聚集。具体来说,aBB树需要线性空间,可以在(O(log _B N))I / O中回答任何查询,并且可以在(O(log _B N))I / O中进行更新(其中(N)是总数)在数据集数量为(O(B))的情况下,所有候选数据集的大小,以及(B)磁盘页面大小。广泛的实验证明了我们技术的实际效率。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号