ABSTRACT
Space and time are indispensable for many objects in the real world. Spatial databases represent, store and manipulate spatial data such as points, lines, areas, surfaces and hypervolumes in multidimensional space. Most of these databases suffer from, what is commonly called, the “Curse of Dimensionality” [1]. Curse of dimensionality refers to a performance degradation of similarity queries with increasing dimensionality of these databases. One way to reduce this curse is to develop data structures for indexing such databases to answer similarity queries efficiently. Specialized data structures such as R-trees and its variants (Chapter 21), have been proposed for this purpose which have demonstrated multi-fold performance gains in access time on this data over sequential search.
