Data declustering for efficient range and similarity searching

S. Prabhakar, D. Agrawal, A. El Abbadi

Research output: Chapter in Book/Report/Conference proceedingConference contribution

4 Citations (Scopus)

Abstract

Advances in processor and network technologies have catalyzed the growth of data intensive applications such as image repositories and digital libraries. The lack of commensurate improvements in storage systems has resulted in I/O becoming a major bottleneck in modern systems. The use of parallel I/O from multiple devices is a well known technique for improving I/O performance. A key factor in exploiting parallel I/O is knowledge of the access pattern - the sets of data items that are likely to be accessed concurrently should be declustered across the disks. Range and nearest-neighbor (similarity) queries are the most important class of queries for multimedia databases. Declustering schemes tailored for improving the performance of range only or similarity only queries have been proposed in the literature. The problem of declustering for combined range and similarity queries has not been addressed in the literature. We evaluate the performance of declustering schemes for combined range and similarity queries. It is established that the Cyclic allocation schemes, that we have developed, give the best overall performance improvement for combined queries and are most robust with respect to variations in system parameters. The evaluation is based upon the ability to achieve parallel I/O. We consider both combined queries as well as independent range and similarity queries with a given declustering.

Original languageEnglish
Title of host publicationProceedings of SPIE - The International Society for Optical Engineering
EditorsC.C. Jay Kuo, S. Fu Chang, S. Panchanathan
Pages419-430
Number of pages12
Volume3527
DOIs
Publication statusPublished - 1998
Externally publishedYes
EventMultimedia Storage and Archiving Systems III - Boston, MA, United States
Duration: 2 Nov 19984 Nov 1998

Other

OtherMultimedia Storage and Archiving Systems III
CountryUnited States
CityBoston, MA
Period2/11/984/11/98

Fingerprint

Digital libraries
multimedia
central processing units
evaluation

Keywords

  • Data Placement
  • Declustering
  • Multimedia Databases
  • Parallel I/O

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Condensed Matter Physics

Cite this

Prabhakar, S., Agrawal, D., & El Abbadi, A. (1998). Data declustering for efficient range and similarity searching. In C. C. Jay Kuo, S. Fu Chang, & S. Panchanathan (Eds.), Proceedings of SPIE - The International Society for Optical Engineering (Vol. 3527, pp. 419-430) https://doi.org/10.1117/12.325834

Data declustering for efficient range and similarity searching. / Prabhakar, S.; Agrawal, D.; El Abbadi, A.

Proceedings of SPIE - The International Society for Optical Engineering. ed. / C.C. Jay Kuo; S. Fu Chang; S. Panchanathan. Vol. 3527 1998. p. 419-430.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Prabhakar, S, Agrawal, D & El Abbadi, A 1998, Data declustering for efficient range and similarity searching. in CC Jay Kuo, S Fu Chang & S Panchanathan (eds), Proceedings of SPIE - The International Society for Optical Engineering. vol. 3527, pp. 419-430, Multimedia Storage and Archiving Systems III, Boston, MA, United States, 2/11/98. https://doi.org/10.1117/12.325834
Prabhakar S, Agrawal D, El Abbadi A. Data declustering for efficient range and similarity searching. In Jay Kuo CC, Fu Chang S, Panchanathan S, editors, Proceedings of SPIE - The International Society for Optical Engineering. Vol. 3527. 1998. p. 419-430 https://doi.org/10.1117/12.325834
Prabhakar, S. ; Agrawal, D. ; El Abbadi, A. / Data declustering for efficient range and similarity searching. Proceedings of SPIE - The International Society for Optical Engineering. editor / C.C. Jay Kuo ; S. Fu Chang ; S. Panchanathan. Vol. 3527 1998. pp. 419-430
@inproceedings{4af9a0e4d8614897951437e81da288ea,
title = "Data declustering for efficient range and similarity searching",
abstract = "Advances in processor and network technologies have catalyzed the growth of data intensive applications such as image repositories and digital libraries. The lack of commensurate improvements in storage systems has resulted in I/O becoming a major bottleneck in modern systems. The use of parallel I/O from multiple devices is a well known technique for improving I/O performance. A key factor in exploiting parallel I/O is knowledge of the access pattern - the sets of data items that are likely to be accessed concurrently should be declustered across the disks. Range and nearest-neighbor (similarity) queries are the most important class of queries for multimedia databases. Declustering schemes tailored for improving the performance of range only or similarity only queries have been proposed in the literature. The problem of declustering for combined range and similarity queries has not been addressed in the literature. We evaluate the performance of declustering schemes for combined range and similarity queries. It is established that the Cyclic allocation schemes, that we have developed, give the best overall performance improvement for combined queries and are most robust with respect to variations in system parameters. The evaluation is based upon the ability to achieve parallel I/O. We consider both combined queries as well as independent range and similarity queries with a given declustering.",
keywords = "Data Placement, Declustering, Multimedia Databases, Parallel I/O",
author = "S. Prabhakar and D. Agrawal and {El Abbadi}, A.",
year = "1998",
doi = "10.1117/12.325834",
language = "English",
volume = "3527",
pages = "419--430",
editor = "{Jay Kuo}, C.C. and {Fu Chang}, S. and S. Panchanathan",
booktitle = "Proceedings of SPIE - The International Society for Optical Engineering",

}

TY - GEN

T1 - Data declustering for efficient range and similarity searching

AU - Prabhakar, S.

AU - Agrawal, D.

AU - El Abbadi, A.

PY - 1998

Y1 - 1998

N2 - Advances in processor and network technologies have catalyzed the growth of data intensive applications such as image repositories and digital libraries. The lack of commensurate improvements in storage systems has resulted in I/O becoming a major bottleneck in modern systems. The use of parallel I/O from multiple devices is a well known technique for improving I/O performance. A key factor in exploiting parallel I/O is knowledge of the access pattern - the sets of data items that are likely to be accessed concurrently should be declustered across the disks. Range and nearest-neighbor (similarity) queries are the most important class of queries for multimedia databases. Declustering schemes tailored for improving the performance of range only or similarity only queries have been proposed in the literature. The problem of declustering for combined range and similarity queries has not been addressed in the literature. We evaluate the performance of declustering schemes for combined range and similarity queries. It is established that the Cyclic allocation schemes, that we have developed, give the best overall performance improvement for combined queries and are most robust with respect to variations in system parameters. The evaluation is based upon the ability to achieve parallel I/O. We consider both combined queries as well as independent range and similarity queries with a given declustering.

AB - Advances in processor and network technologies have catalyzed the growth of data intensive applications such as image repositories and digital libraries. The lack of commensurate improvements in storage systems has resulted in I/O becoming a major bottleneck in modern systems. The use of parallel I/O from multiple devices is a well known technique for improving I/O performance. A key factor in exploiting parallel I/O is knowledge of the access pattern - the sets of data items that are likely to be accessed concurrently should be declustered across the disks. Range and nearest-neighbor (similarity) queries are the most important class of queries for multimedia databases. Declustering schemes tailored for improving the performance of range only or similarity only queries have been proposed in the literature. The problem of declustering for combined range and similarity queries has not been addressed in the literature. We evaluate the performance of declustering schemes for combined range and similarity queries. It is established that the Cyclic allocation schemes, that we have developed, give the best overall performance improvement for combined queries and are most robust with respect to variations in system parameters. The evaluation is based upon the ability to achieve parallel I/O. We consider both combined queries as well as independent range and similarity queries with a given declustering.

KW - Data Placement

KW - Declustering

KW - Multimedia Databases

KW - Parallel I/O

UR - http://www.scopus.com/inward/record.url?scp=0038410109&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0038410109&partnerID=8YFLogxK

U2 - 10.1117/12.325834

DO - 10.1117/12.325834

M3 - Conference contribution

VL - 3527

SP - 419

EP - 430

BT - Proceedings of SPIE - The International Society for Optical Engineering

A2 - Jay Kuo, C.C.

A2 - Fu Chang, S.

A2 - Panchanathan, S.

ER -