An efficient and interactive A*-algorithm with pruning power: materialized view selection revisited

摘要

Materialized view selection with resource constraint is one of the most important issues in the management of data warehouses. In this paper, we revisit the problem of materialized view selection under disk-space constraint S. Many efficient greedy algorithms have been developed. However, we observe that when S is small, their solution quality will not be well guaranteed. In order to further improve solution quality in such cases, we develop a competitive A* algorithm. Both theory and experiment results show that our algorithm is a powerful, efficient and flexible scheme for this problem.