Adaptive maintenance for classified sliding window joins over data streams

作者：Gu Yu; Yue Dejun; Li Xiaojing; Wu Shanshan; Lv Yanfei; Yu Ge

来源：Journal of Computational Information Systems, 2007, 3(5): 1809-1814.

• Data form
• Data stream
• Join
• Sliding window
• Time based windows

As a new kind of data form different from data stored in traditional database, data stream is being applied in more and more areas, in which sliding window joins are used frequently and incur high processing cost The existing maintenance strategies for sliding windows joins focus more on time-based windows rather than tuple-based windows, partitioned windows and other window types which are also important in complex applications. In this paper, join models and relative strategies over data stream are analyzed and an optimized method named AT based on negative tuple method is proposed. Furthermore, an adaptive strategy CWJ is discussed in which join strategies may be adjusted depending on the various parameters. The extensive experiments with synthetic and real data display the efficiency and flexibility of our strategy for classified sliding window joins.

• 出版日期2007
• 单位东北大学