Efficient high-dimensional indexing by superimposing space-partitioning schemes

The problem of accessing data in high-dimensional spaces has attracted considerable scientific interest. Multidimensional access methods typically employ a space-partitioning strategy whose goal is to direct the search toward relevant parts of the space. Unfortunately, monolithic partitioning, whether it is static, dynamic or some hybrid scheme, has serious limitations in high-dimensional situations. This work presents a new family of indexing techniques that superimposes static partitioning over a dynamic partitioning scheme in order to improve the search performance in high-dimensional spaces. A comprehensive set of experiments shows that, in realistic scenarios, the proposed solutions are usually much more efficient and more scalable than the multidimensional access methods based on monolithic partitioning.