关键词: 大规模数据集, Top-k Skyline, 代表性信息, 并行化处理, 筛选规则

Abstract: As data of an unprecedented scale are becoming accessible, it becomes more and more important to help user identify the representative information of a manageable size. Top-k Skyline queries can find the most k representative information, and it can also control the size of the results. So the Top-k Skyline queries meet the above-mentioned requirements. However, the traditional method of Top-k Skyline query has a low efficiency when it meets a large scale of data set. In order to speed the efficiency of the query, parallel processing comes into being. Via combining the Top-k Skyline query with the parallel processing, this paper proposes a novel algorithm of parallel Top-k Skyline queries for large scale of data set, named PTKS (parallel Top-K Skyline). With exploiting the advantage of parallel computing, making the efficient query by using parallel processing, and designing a filter rule based on user preferences to reduce data size, this algorithm satisfies the needs of users on some aspects. Through the experiments on the factual data sets, compared with the existing methods, the PTKS can be applied to the large scale of data set and is superior to the other existing algorithms in large data set.

Key words: large data set, Top-k Skyline, representative information, parallel processing, filter rule