Paper Info
Title
Categorized Sliding Window in Streaming Data Management Systems
Abstract
For many applications, data is collected at very large rates from various sources. Applications that produce results from this data have a requirement for very efficient processing in order to achieve timely decisions. An example of such a demanding applications is one that takes decisions on stock acquisition based on the price updates that happen constantly while the market is open for transactions. Our proposed technique is a simple yet effective way to reduce the access time to the streaming data.In this paper we propose an efficient indexing technique that improves the access time to data elements in sliding windows of streamed database systems. This technique, called Categorized Sliding Window, is based on splitting the data into categories and using bit vectors to avoid accesses to non-relevant data.Our experimental results show large improvements compared with simpler techniques. For the standard Linear Road benchmark we observe a performance improvement of 3.3x for a complex continuous query. Also relevant is the fact that 90% of the performance improvement is achieved with only 65% of the maximum number of categories, which represents a memory overhead of only 13.5%.
Year
DOI
Venue
2008
10.1007/978-3-540-85654-2_52
DEXA
Keywords
Field
DocType
performance improvement,large improvement,efficient processing,access time,categorized sliding window,proposed technique,data element,streaming data management systems,efficient indexing technique,large rate,simpler technique,bit vector,indexation,database system,sliding window
Data mining,Sliding window protocol,Access time,Data element,Computer science,Search engine indexing,Real-time computing,Streaming data,Management system,Database,Performance improvement
Conference
Volume
ISSN
Citations 
5181
0302-9743
0
PageRank 
References 
Authors
0.34
16
3
Name
Order
Citations
PageRank
Marios Papas11197.27
Josep-L. Larriba-Pey216217.44
Pedro Trancoso337743.79