Efficient Skyline Computation of Multiple Range Skyline Queries
Zarina Binti Dzolkhifli, Hamidah Ibrahim, Fatimah Sidi, Lilly Suriani Affendey, Siti Nurulain Mohd Rum, Ali Amer Alwan · 2021
Skyline query which returns a set of skyline objects by filtering those objects that are dominated by others from a potentially large multidimensional data set has attracted significant research attention especially in the database community. Many variants of skyline queries have been introduced which include among others range skyline queries which retrieve skyline objects within a specified range also known as condition (constraint). It enables users to specify preferences within an ideal range value over a dimension(s) instead of a single sought value. Although range skyline queries have been studied extensively, most of the works focus on the optimisation problem of skyline computation for a given range skyline query. Nonetheless, deriving skyline objects for multiple range skyline queries separately is unwise since these queries might specify similar constraints/sub constraints and hence the same set of objects has to be scanned and compared multiple times before the skyline objects can be determined for each individual query. In this paper, we propose the Reduct-Sky framework that attempts to avoid unnecessary skyline computations of multiple range skyline queries. Intuitively, this is achieved by analysing the conditions of multiple range skyline queries to mainly extract the conditions/sub conditions that are the same to ensure that skyline computation over the involved set of objects is performed only once instead of multiple times. An initial performance evaluation of the proposed framework demonstrates its efficiency with regard to number of pairwise comparisons.