Show simple item record

Large Scale Disk-Based Metric Indexing Structure Approximate Information Retrieval by Content

dc.contributor.authorRukoz, Marta
dc.contributor.authorGouet-Brunet, Valérie
HAL ID: 170517
ORCID: 0000-0003-3666-5146
dc.contributor.authorBarton, Stanislav
dc.date.accessioned2011-06-20T12:19:52Z
dc.date.available2011-06-20T12:19:52Z
dc.date.issued2011
dc.identifier.urihttps://basepub.dauphine.fr/handle/123456789/6540
dc.language.isoenen
dc.subjectApproximate Information Retrievalen
dc.subjectlarge scalabilityen
dc.subjectindex structuresen
dc.subject.ddc005.7en
dc.titleLarge Scale Disk-Based Metric Indexing Structure Approximate Information Retrieval by Contenten
dc.typeCommunication / Conférence
dc.contributor.editoruniversityotherPOND University;France
dc.contributor.editoruniversityotherCNAM/CEDRIC;France
dc.description.abstractenIn order to achieve large scalability, indexing structures are usually distributed to incorporate more of expensive main memory during the query processing. In this paper, an in- dexing structure, that does not su er from a performance degradation by its transition from main memory storage to hard drive, is proposed. The high e ciency of the index is achieved using a very e ective pruning based on precom- puted distances and so called locality phenomenon which substantially diminishes the number of retrieved candidates. The trade-o s for the large scalability are, rstly, the ap- proximation and, secondly, longer query times, yet both are still bearable enough for recent multimedia content-based search systems, proved by an evaluation using visual and audio data and both metric and semi-metric distance func- tions. The tuning of the index's parameters based on the analysis of the particular's data intrinsic dimensionality is also discussed.en
dc.identifier.citationpages2-7en
dc.relation.ispartoftitleEDBT/ICDT '11en
dc.relation.ispartofeditorStefanova, Silvia
dc.relation.ispartofeditorOrsborn, Kjell
dc.relation.ispartofeditorDeepak, P
dc.relation.ispartofeditorDeshpande, Prasad
dc.relation.ispartofpublnameACMen
dc.relation.ispartofpublcityNew Yorken
dc.relation.ispartofdate2011
dc.relation.ispartofpages36en
dc.description.sponsorshipprivateouien
dc.subject.ddclabelOrganisation des donnéesen
dc.relation.ispartofisbn978-1-4503-0612-6en
dc.relation.conftitle1st Workshop on New Trends in Similarity Search (NTSS’11), in conjunction with the EDBT 2011 Confereen
dc.relation.confdate2011-03
dc.relation.confcityUppsalaen
dc.relation.confcountrySuèdeen


Files in this item

FilesSizeFormatView

There are no files associated with this item.

This item appears in the following Collection(s)

Show simple item record