Parallelization of group‐based skyline computation for multi‐core processors

摘要: Summary Skyline computation is particularly useful in multi-criteria decision-making applications. However, it inadequate to answer queries that need analyze not only individual points but also groups of points. Compared the traditional skyline computation, computing group-based much more complicated and expensive. This computational challenge promotes us use modern platforms accelerate computation. In this paper, we introduce a novel multi-core algorithm compute skyline. We first layers data set parallel, which are critical intermediate result. algorithm, maintain an efficiently updatable structure for shared global layers, used minimize dominance tests high throughput. Then design efficient parallel find based on layers. Extensive experimental results real synthetic sets show our algorithms achieve 10-fold speedup with 16 threads over state-of-the-art sequential challenging workloads.