中文摘要：

随着维度和数据量的不断增长,已有的轮廓查询方法已无法为用户提供更好的查询支持.本文采用并行处理框架MapReduce 结合k-支配轮廓查询算法来解决此类查询问题.在对输入的数据集进行预排序的基础上,提出了基于MapReduce 的单轮扫描算法和双轮扫描算法,通过在不同维度、数据分布、节点个数等方面的实验测试及对比,验证了算法的有效性.

英文摘要：

With the increasing dimension and dataset, the existing skyline algorithm cannot support better decision for users. Inthis paper, the k-dominant skyline algorithm combined with MapReduce framewok is used to slove the skyline query problem.Based on presorting the input datasets to optimized execution,two algorithms,MapReduce based one scan algorithmandMapReducebased two scan algorithm, are proposed. Extensive experiments to evaluate two algorithms on different dimensions, data distributionand number of cluster are implemented, and the effectiveness of two algorithms are proved.