Paper Info
Title
A Structure for Sliding Window Equijoins in Data Stream Processing
Abstract
Sliding window equijoins are commonly used in data stream applications. In their implementation, a hash table is generally allocated for each stream source. However, this structure may degrade join performance because all tuples in the hash tables need to be scanned to find expired tuples whenever windows are updated periodically. This paper presents an alternative structure to improve the performance of sliding window equijoins. The basic idea of the proposed method is to allocate a hash table for each set of tuples arriving for a window slide interval. This organization facilitates the identification of expired tuples and improves the join performance, which we observed through our experimental results.
Year
DOI
Venue
2013
10.1109/CSE.2013.25
C3S2E
Keywords
DocType
ISSN
alternative structure,hash table allocation,sliding window equijoins,data streams,windowing structure,window equijoins,window joins,expired tuples identification,expired tuples,basic idea,data stream application,file organisation,hash table,window slide interval,data stream processing,data mining,symmetric hash joins,stream source,sliding windows
Conference
1949-0828
Citations 
PageRank 
References 
2
0.36
11
Authors
1
Name
Order
Citations
PageRank
Hyeon Gyu Kim1145.03