【作者】 黄炎灿； 董玮； 陈立军；

【机构】 北京大学计算机科学技术系；

【摘要】 <正>1引言近几年来,数据流上的查询处理备受关注。与描述稳定数据集的关系数据模型不同,数据流模型描述的是不断到达的元组序列,它具有实时、连续、有序、无界的特征,其流速也随时间而变化。引更多还原

【Abstract】 Join operator plays an important role in data stream query processing.Join operator on data stream resembles that in relational algebra,except for introducing the window restriction.Due to the real-time,continuous,ordered and unlimited property of data streams,join operator is demanded to output as many results as possible,continually and smoothly,and to be self-adaptive to bursty input streams.In this paper,we propose a join algorithm, WJoin,which is based on Symmetric Hash Table and is proper for incremental computation.Experiment studies show that WJoin is an effective solution for providing good responses to users in the environment of data stream.更多还原