Computing Partial Data Cubes.

Frank Dehne, Todd Eavis, and Andrew Rau-Chaplin

Abstract: The precomputation of the different views of a data cube is critical to improving the response time of data cube queries for On-Line Analytical Processing (OLAP). However, the user is often not interested in the set of all views of the data cube but only in a certain subset of views. In this paper, we study the problem of computing the partial data cube, i.e. a subset of selected views in the lattice. We consider the case of dense relations, using top-down cube construction methods like Pipesort. This paper presents, both, sequential and parallel methods for partial data cube construction as well as an experimental performance evaluation of our methods.