Storage technology of high-dimensional OLAP aggregate data in data warehouse system

For the cube with d dimensions,it can generate 2d cuboids.But in a high-dimensional cube,it might not be practical to build all these cuboids.To deal with this problem,a novel approach for Online Analytical Processing(OLAP) in high-dimensional datasets by partitioning the high dimensional cube into some low-dimensional shell segment mini-Cubes was proposed.OLAP queries were computed online by dynamically constructing cuboids from these shell segment mini-cubes through the parallel and distributed processing system.With this design,for high-dimensional OLAP,the total space that needs to store such shell segment mini-cubes was negligible in comparison with a high-dimensional cube.The methods of shell mini-cube was compared to other existed ones such as full cube and partial cube by experiment.The analytical and experimental results showed that the algorithms of segment mini-cubes proposed were more efficient than the other existing ones.