Given an OLAP query expressed over a collection of source OLAP databases, we study the problem of computing the result OLAP target database when the underlying detail data of the source databases are not available. The problem arises when the source databases do not have the same category attributes, and it is not possible to derive the result from a single database. The method we use is the linear indirect estimator. We consider two obvious methods for computing such a target database, called the Full-cross-product (F) and the Pre-aggregation (P) methods. As for the accuracy and computational performance of these methods, while the method F provides the most accurate estimate possible, it is more expensive computationally than P. Our contribution is in proposing a third new method, called the Shortcut method (S), which is less expensive than F, but is as accurate.
展开▼
