Efficient Computation of Iceberg Quotient Cube by Bounding

Quotient cube is a summary structure for a data cube that preserves its semantics. The iceberg cubing problem is to compute the multidimensional group-by partitions that satisfy given aggregation constraints. As we know, there has been no algorithm that computes iceberg quotient cube for nonantimonotone aggregate functions. In this paper, we propose a new structure hyper-star-tree and an efficient algorithm, called IQ-Cubing, for iceberg quotient cubing with nonantimonotone aggregation constraints. We also employ the closedness measure to do pruning efficiently and utilize the closed mask to help the formation of equivalence classes. We conduct an investigation into the performance of our ideas and techniques.