Parallel Computation Of Skyline Queries Over Uncertain Data Based On Multicore Architectures

With the development of data acquisition and processing technology, the uncertainty of data has gradually gotten more attention. In many real-life applications, due to the complex external factors, the data collected from the randomness, incompleteness and uncertainty can truly reflect the status of the real world. Therefore, the management and query of uncertain data has become a hot research topic in recent years. It has some theoretical significance and application value. This paper aims at uncertain data over Multi-Core;(a research problem of how to implement an efficient skyline computation). To address this concern, the main contents are as follows:First, we studied the current classical algorithms of Skyline queries, mainly analyzed implementation idea of probability skyline and distributed Skyline query algorithm. Taking into account the advantages and disadvantages of existing algorithms, we proposed the basic idea of parallel Skyline query over ??uncertain data on multi-core architecture.Secondly, we proposed a parallel computation of skyline queries over uncertain data on multi-core, applied to high-dimensional and skyline computation of large-scale dataset. Based on skyline set computation and probability skyline computation methods, we proposed PSUDM algorithm, and through an example of Baseball players of Chicago White Sox, we described the process of our algorithm.Finally, we used C ++ programming and OPENMP compiler to implement PSUDM algorithm, and evaluate PSUDM algorithm uncertain data sets respectively through different dimensional values, different large-scale of dataset values and different numbers of cores. For further explanation, we implemented Sequential Skyline computation over Uncertain Data on Multi-core architecture(SSUDM),compared execution time of Sequential and Parallel Skyline Queries methods from different dimensional values and different sizes of data sets respectively.Experimental results show that our algorithm successfully exploits features of multi-core to improve the performance of parallel Skyline computation with uncertain database.