In a Multiple Attribute Tree (MAT) based data organization, the average case response to a specific range query depends on the structural properties of MAT. These structural properties depend very much on the interrel...
详细信息
In a Multiple Attribute Tree (MAT) based data organization, the average case response to a specific range query depends on the structural properties of MAT. These structural properties depend very much on the interrelationships among the data elements. Efficiency in searching can be achieved by exploiting the data properties in the construction of MAT. The order or ranking of attributes is a key factor in deciding the profile of the MAT for given data. In this paper, we estimate the average cost of a range query in MAT based data organization. We then prove that the average performance can be improved by ranking the attributes in such a way that the average size of the filial sets decreases towards the lower levels of the tree structure.
暂无评论