Enhancing Similarity Search Throughput by Dynamic Query
Reordering

D 2016

Enhancing Similarity Search Throughput by Dynamic Query Reordering

NÁLEPA, Filip, Michal BATKO and Pavel ZEZULA

Basic information

Original name

Enhancing Similarity Search Throughput by Dynamic Query Reordering

Authors

NÁLEPA, Filip (203 Czech Republic, guarantor, belonging to the institution), Michal BATKO (203 Czech Republic, belonging to the institution) and Pavel ZEZULA (203 Czech Republic, belonging to the institution)

Edition

Cham, Database and Expert Systems Applications: 27th International Conference, DEXA 2016, Porto, Portugal, September 5-8, 2016, Proceedings, Part II, p. 185-200, 16 pp. 2016

Publisher

Springer International Publishing

Other information

Language

English

Type of outcome

Stať ve sborníku

Field of Study

10201 Computer sciences, information science, bioinformatics

Country of publisher

Portugal

Confidentiality degree

není předmětem státního či obchodního tajemství

Publication form

printed version "print"

Impact factor

Impact factor: 0.402 in 2005

RIV identification code

RIV/00216224:14330/16:00088102

Organization unit

Faculty of Informatics

ISBN

978-3-319-44405-5

ISSN

DOI

http://dx.doi.org/10.1007/978-3-319-44406-2_14

UT WoS

000389020200014

Keywords in English

Stream processing; Similarity Search

Abstract

V originále

A lot of multimedia data are being created nowadays, which can only be searched by content since no searching metadata are available for them. To make the content search efficient, similarity indexing structures based on the metric-space model can be used. In our work, we focus on a scenario where the similarity search is used in the context of stream processing. In particular, there is a potentially infinite sequence (stream) of query objects, and a query needs to be executed for each of them. The goal is to maximize the throughput of processed queries while maintaining an acceptable delay. We propose an approach based on dynamic reordering of the incoming queries combined with caching of recent results. We were able to achieve up to 3.7 times higher throughput compared to the base case when no reordering and caching is used.

Links

GA16-18889S, research and development project

Name: Analytika pro velká nestrukturovaná data (Acronym: Big Data Analytics for Unstructured Data)

Investor: Czech Science Foundation

Citovat

NÁLEPA, Filip, Michal BATKO and Pavel ZEZULA. Enhancing Similarity Search Throughput by Dynamic Query Reordering. In Hartmann, Sven and Ma, Hui. Database and Expert Systems Applications: 27th International Conference, DEXA 2016, Porto, Portugal, September 5-8, 2016, Proceedings, Part II. Cham: Springer International Publishing, 2016, p. 185-200. ISBN 978-3-319-44405-5. Available from: https://dx.doi.org/10.1007/978-3-319-44406-2_14.

@inproceedings{1352385,
   author = {Nálepa, Filip and Batko, Michal and Zezula, Pavel},
   address = {Cham},
   booktitle = {Database and Expert Systems Applications: 27th International Conference, DEXA 2016, Porto, Portugal, September 5-8, 2016, Proceedings, Part II},
   doi = {http://dx.doi.org/10.1007/978-3-319-44406-2_14},
   editor = {Hartmann, Sven and Ma, Hui},
   keywords = {Stream processing; Similarity Search},
   howpublished = {tištěná verze "print"},
   language = {eng},
   location = {Cham},
   isbn = {978-3-319-44405-5},
   pages = {185-200},
   publisher = {Springer International Publishing},
   title = {Enhancing Similarity Search Throughput by Dynamic Query Reordering},
   year = {2016}
}

TY  - JOUR
ID  - 1352385
AU  - Nálepa, Filip - Batko, Michal - Zezula, Pavel
PY  - 2016
TI  - Enhancing Similarity Search Throughput by Dynamic Query Reordering
PB  - Springer International Publishing
CY  - Cham
SN  - 9783319444055
KW  - Stream processing
KW  - Similarity Search
N2  - A lot of multimedia data are being created nowadays, which can only be searched by content since no searching metadata are available for them. To make the content search efficient, similarity indexing structures based on the metric-space model can be used. In our work, we focus on a scenario where the similarity search is used in the context of stream processing. In particular, there is a potentially infinite sequence (stream) of query objects, and a query needs to be executed for each of them. The goal is to maximize the throughput of processed queries while maintaining an acceptable delay. We propose an approach based on dynamic reordering of the incoming queries combined with caching of recent results. We were able to achieve up to 3.7 times higher throughput compared to the base case when no reordering and caching is used.
ER  -

NÁLEPA, Filip, Michal BATKO and Pavel ZEZULA. Enhancing Similarity Search Throughput by Dynamic Query Reordering. In Hartmann, Sven and Ma, Hui. \textit{Database and Expert Systems Applications: 27th International Conference, DEXA 2016, Porto, Portugal, September 5-8, 2016, Proceedings, Part II}. Cham: Springer International Publishing, 2016, p.~185-200. ISBN~978-3-319-44405-5. Available from: https://dx.doi.org/10.1007/978-3-319-44406-2\_{}14.

Detailed Information on Publication Record