PRS: efficient range skyline computation on massive data via presorting

摘要

In many applications, range skyline query is an important operation to find the interesting tuples in a potentially huge data space. Given selection condition, range skyline query returns tuples both satisfying the specified selection condition and not dominated by other satisfying tuples. It is found that most of the existing skyline algorithms do not consider the selection condition. This paper proposes a novel table-scan-based Presorted-table-based Range Skyline (PRS) algorithm to efficiently compute range skyline results on massive data. PRS first presorts the table for early termination. The early termination checking is proposed in this paper, along with the theoretical analysis of scan depth. The selection checking and dominance checking are devised in this paper to skip the unsatisfying or dominated tuples directly. The theoretical analysis proves that the overwhelming majority of candidates can be skipped. The extensive experimental results, conducted on synthetic and real-life data sets, show that PRS outperforms the existing algorithms significantly.