摘要
目前对数据流连续查询的研究工作大部分基于滑动窗口,窗口大小固定不变.本文从内存资源有限性与查询结果精确性角度出发,采用可变窗口技术,在缓存中加入窗口控制器,当数据到达缓存时,根据用户提出查询具体问题和数据流速情况,窗口分配算子对其进行窗口大小的划分.使得在内存资源有限的情况下,通过动态改变滑动窗口大小,充分利用系统资源,降低内存使用率的同时,使连续查询的查询结果都在允许的误差范围内,此方法经过验证是可行的.
At present, the research work which inquires data stream continuously is all based on a sliding- window whose size is invariable. This article embarks from the finiteness of the memory resources and the veracity of the inquiry result using the variable window technology. The window controller was put in the buffer. When the data arrived, the window was changed dynamically according to the user proposed question and the speed of data stream flow. The window assignment - operators carry on the division of the window to make the best use of the system re- sources and reduce the utilization ratio of the EMS memory. Contemporary, the inquired results are all in the permis- sion erroneous scope. The feasibility of this method is proved through test.
出处
《佳木斯大学学报（自然科学版）》
CAS
2009年第5期694-697,共4页
Journal of Jiamusi University：Natural Science Edition
基金
黑龙江省教育厅科学技术研究项目资助(11521008)
关键词
数据流
滑动窗口
连续查询
窗口分配算子
data stream
sliding-window
continuous query
window assignment-operator
作者简介
作者简介：吴亚娟（1983-），女，硕士研究生，主要从事数据流挖掘方面的研究．
