Efficiently compressing OLAP data cubes via R-tree based recursive partitions

This paper extends a quad-tree based multi-resolution approach for two-dimensional summary data by providing a novel OLAP data cube compression approach that keeps the critical novelty of relying on R-tree based partitions instead of more constrained classical kinds of partition. This important novelty introduces the nice amenity of (i) allowing end-users to exploit the semantics of data and (ii) obtaining compressed representations of data cubes where more space can be invested to describe those ranges of multidimensional data for which they retain a higher degree of interest. Hence, this paper can be considered as an important advancement over the state-of-the-art approaches. Experimental results confirm the benefits of our proposed approach. © 2012 Springer-Verlag.