دانلود رایگان مقاله رویکرد کارآمد برای کوئری اسکای لاین روی داده نامشخص توزیع شده

Abstract

The skyline query as an important aspect of big data management, has received considerable attention from the database community, due to its importance in many applications including multi-criteria decision making, preference answering, and so forth. Moreover, the uncertain data from many applications have become increasing distributed, which makes the central assembly of data at one location for storage and query infeasible and inefficient. The lack of global knowledge and the computational complexity derived from the introduction of the data uncertainty make the skyline query over distributed uncertain data extremely challenging. Although many efforts have addressed the skyline query problem over various distributed scenarios, existing studies still lack the approaches to efficiently process the query. In this paper, we extensively study the distributed probabilistic skyline query problem and propose an efficient approach GDPS to address the problem with an optimized iterative feedback mechanism based on the grid summary. Furthermore, many strategies for further optimizing the query are also proposed, including the optimization strategies for the local pruning, tuple selecting and the server pruning. Extensive experiments on real and synthetic data sets have been conducted to verify the effectiveness and efficiency of our approach by comparing with the state-of-the-art approaches.

8. Conclusions

In this paper, we have addressed the problem of skyline queries over distributed uncertain data sets. To accelerate the query pro-cessing, we propose an efficient pruning mechanism for preprocessing with grid summary. Furthermore, we propose many strategies for optimizing the queries based on a feedback mechanism. Extensive experimental results with real data and synthetic data have verified the effectiveness and efficiency of the proposals. In our future work, we will consider querying the skylines over complex distributed uncertain data streams, as there are many potential demands for continuous skyline query currently [53].