用于聚集值近似查询的基于密度的树索引结构

A Density-based Tree Structure for Approximating Aggregate Queries

如何快速有效地对数据立方体上的聚集查询给出近似的回答,是数据挖掘和数据仓库研究领域中的核心问题之一。现有大多数聚集查询算法在同一个数据立方体上只能支持某种特定的而非多种类型的聚集查询。本文给出了一种新的框架AdenTS,即基于密度的自适应树结构,它可以回答同一数据立方体上的各类聚集查询,也提出了一些近似和启发式技术,改善了查询结果和精度。实验结果表明,这种方法在支持的查询种类和性能上是更好的。

In many fields and applications, it is critical for users to make decisions through OLAP queries. How are accuracy and efficiency promoted while answering multiple aggregate queries, e.g. COUNT, SUM, AVG, MAX, MIN and MEDIAN？ It has been the urgent problem in the fields of OLAP and data summarization recently. There have been a few solutions such as MRA-Tree and GENHIST for it. However, they could only answer a certain aggregate query which was defined in a particular data cube with some limited applications. In this paper, we develop a novel framework ADenTS, e.g. Adaptive Density-based Tree Structure, to answer various types of aggregate queries within a single da- ta cube. We represent the whole cube by building a coherent tree structure. Furthermore, several techniques for approximation and heuristic approaches are proposed to improve the accuracy of query answering. The experimental results show that the method outperforms others in effectiveness and efficiency.