Range-max/min query in OLAP data cube

Abstract

A range query applies an aggregation operation over all selected cells of an OLAP data cube where the selection is specified by ranges of continuous values for numeric dimensions. Much work has been done with one type of aggregations: SUM. But little work has been done with another type of aggregations: MAX/MIN besides the tree-based algorithm. In this paper, we propose a new method which partitions the given data cube, stores precomputed max/min over partitions and location of the max/min of the partitions. We also use some techniques to reduce the chance of accessing the original data cube when answering ad hoc queries at run time. The experiment results demonstrate that our method outperforms the tree-based algorithm.