An Empirical Comparison of Methods for Iceberg-CUBE Construction

The Iceberg-Cube problem is to apply an aggregate function over a set of attributes to determine which combinations of attribute values are above a specified aggregate threshold. We implemented bottom-up and top-down methods for this problem. The bottom-down method we used already used pruning. Results show that even when the top-down method employed pruning, it was slower than the bottom-up method because it did not prune as efficiently.