面向高速乱序流的top-k连续查询方法  

Continuous Top-k Query Method over High-speed Out-of-order Data Streams

作　　者：武守晓 房俊[1,2] WU Shouxiao;FANG Jun(Beijing Key Laboratory on Integration and Analysis of Large-scale Stream Data,North China University of Technology,Beijing 100144,China;Institute of Data Engineering,North China University of Technology,Beijing 100144,China)

机构地区：[1]北方工业大学大规模流数据集成与分析技术北京市重点实验室,北京100144 [2]北方工业大学数据工程研究院,北京100144

出　　处：《郑州大学学报（理学版）》2021年第3期93-99,共7页Journal of Zhengzhou University:Natural Science Edition

基　　金：国家重点研发计划项目(2017YFC0804406);国家自然科学基金项目(61672042)。

摘　　要：提出一种面向高速乱序流的top-k连续查询方法。使用基于缓存的方法等待迟到元组,但不对缓冲区内数据进行排序,通过统计运行信息实现缓存时长自适应,然后使用改造的MinTopk算法计算当前窗口的top-k结果集。实验结果表明,该方法在高速乱序流上实现了高效的top-k查询,在保证用户允许的最小正确率的情况下计算出最小缓存时长,减少了查询时延。The continuous top-k query approach over high-speed out-of-order data streams was proposed.Using a cache-based method to wait for late tuples without sorting the data in the buffer,the self-adaptive cache duration was realized by counting the running information.And the modified MinTopk algorithm was used to calculate the top-k result set of the current window.The experimental results showed that this approach could achieve efficient top-k query over high-speed out-of-order data streams.In case of ensuring the minimum accuracy allowed by users,the minimum cache duration was calculated to reduce the query delay.

关 键 词：高速乱序流 top-k连续查询 缓存时长自适应 查询时延 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]