An implementation scheme of multidimensional arrays for MOLAP
T. Tsuji, A. Isshiki, Teruhisa Hochin, K. Higuchi · 2004
Multidimensional arrays used in MOLAP are often sparse. They also suffer from the problem that the time consumed in sequential access to array elements heavily depends on the dimension along which the elements are accessed. This problem would be alleviated by dividing the whole array into the set of the same sized smaller subarrays called "chunks". These chunks are also sparse and they are compressed. However these compressed chunks are usually arranged in the predetermined order of the dimensions. This would produce further dimension dependency in accessing array elements. In this paper, by introducing the notion of a "chunk container", we resolve the problem of this further dimension dependency.