一种弹性可扩展的并行n-of-N Skyline查询处理算法
n Elastic and Scalable Algorithm of Parallel n-of-N Skyline Query Processing
n-of-N Skyline查询关注于大小为N的滑动窗口上最近任意n(n<=N)个数据对象的Skyline查询结果，为用户的Skyline查询提供了高度的灵活性。在当前大数据的新环境下，数据流呈现出以下两个特征：一是数据流实时大规模高速到达。二是数据流工作负载的急剧变化性。这对n-of-N Skyline查询的实时性以及自适应扩展提出了更高的要求。而目前针对n-of-N Skyline查询相关的研究都是关注于单机环境下的集中式查询算法，难以同时满足当前新环境下查询的实时性与自适应扩展需求。为此，提出了一种弹性并行查询模型EPM，并且基于EPM模型提出了一种弹性可扩展的并行n-of-N Skyline查询算法(elastic parallel n-of-N Skylline,EPnNS)。实验证明，该算法在并行节点数增加1倍时，查询效率提升接近70%，而且在应对不同程度的负载变化时，该算法具有较好的自适应调整性能。
n-of-N Skyline query processing provides high degree of flexibility for user as it focus on computing the skyline for the most recent n(n≤N) elements in the sliding window of N. Under the new current condition of the big data，the data stream query system presents the following two features. The first is the massive real-time arrival of data. The second is the high uncertainty of workload. And stream with these two features stream demands higher requirements of real-time queries and elastic configurations. Nevertheless, existing studies on n-of-N Skyline query processing only focus on centralized queries in the single-machine environment, which cannot meet the requirement of higher real-time processing and elastic configurations under the new condition of big data. Motivated by the above facts, this paper proposes a elastic parallel model first. Based on this model ,we design a novel elastic and scalable method of Parallel n-of-N Skyline Query Processing(EPnNS). Massive experiments demonstrate that the proposed algorithm performs well over different situations.
马行空、魏炜、王媛、王意洁
计算技术、计算机技术
计算机应用n-of-N Skyline并行弹性数据流处理
omputer Applicationsn-of- N SkylineParallelElasticata stream processing
马行空,魏炜,王媛,王意洁.一种弹性可扩展的并行n-of-N Skyline查询处理算法[EB/OL].(2015-09-23)[2025-08-02].http://www.paper.edu.cn/releasepaper/content/201509-210.点此复制
评论