A Partitioned-Based Method of Convex Skyline for Efficient Processing Top-k Queries

Abstract

A top-k query returns k tuples with the highest (or the lowest) scores from a relation. Layer-based methods are the representative one for processing the top-k query. They construct i-th layer with the objects which can be the top-i, and answer the top-k queries by reading at most k layers. To construct layers, the existing methods used convex skyline, convex hull or skyline. The convex skyline is computed by computing the convex hull over the skyline. Accordingly, the layer size of the convex skyline is relatively smaller than those of the convex hull or the skyline. However, for a high-dimensional databases environment, since most objects can be the skyline points, the convex skyline suffers from a long computing time and a large memory usage. In this paper, we propose a partitioned-based method of convex skyline (pCVX), which reduces the computing time and the memory usage of the convex skyline. We partition the region of the skyline into multiple sub regions and combine the convex hulls, which are computed over sub regions.