Utilize este identificador para referenciar este registo: https://hdl.handle.net/1822/90004

Registo completo
Campo DCValorIdioma
dc.contributor.authorAlmeida, Paulo Sérgiopor
dc.date.accessioned2024-03-25T19:27:28Z-
dc.date.available2024-03-25T19:27:28Z-
dc.date.issued2023-
dc.identifier.citationAlmeida, P. S. (2023). A Case for Partitioned Bloom Filters. IEEE Transactions on Computers. Institute of Electrical and Electronics Engineers (IEEE). http://doi.org/10.1109/tc.2022.3218995por
dc.identifier.issn0018-9340-
dc.identifier.urihttps://hdl.handle.net/1822/90004-
dc.description.abstractIn a partitioned Bloom Filter (PBF) the bit vector is split into disjoint parts, one per hash function. Contrary to hardware designs, where they prevail, software implementations mostly ignore PBFs, considering them worse than standard Bloom filters (SBF), due to the slightly larger false positive rate (FPR). In this paper, by performing an in-depth analysis, first we show that the FPR advantage of SBFs is smaller than thought; more importantly, by deriving the per-element FPR, we show that SBFs have weak spots in the domain: elements that test as false positives much more frequently than expected. This is relevant in scenarios where an element is tested against many filters. Moreover, SBFs are prone to exhibit extremely weak spots if naive double hashing is used, something occurring in mainstream libraries. PBFs exhibit a uniform distribution of the FPR over the domain, with no weak spots, even using naive double hashing. Finally, we survey scenarios beyond set membership testing, identifying many advantages of having disjoint parts, in designs using SIMD techniques, for filter size reduction, test of set disjointness, and duplicate detection in streams. PBFs are better, and should replace SBFs, in general purpose libraries and as the base for novel designs.por
dc.description.sponsorship- (undefined)por
dc.language.isoengpor
dc.publisherIEEEpor
dc.rightsopenAccesspor
dc.subjectInformation filteringpor
dc.subjectpartitioned bloom filterspor
dc.subjectprobabilistic data structurespor
dc.titleA case for partitioned Bloom Filterspor
dc.typearticlepor
dc.peerreviewedyespor
dc.relation.publisherversionhttps://ieeexplore.ieee.org/document/9935317por
oaire.citationStartPage1681por
oaire.citationEndPage1691por
oaire.citationIssue6por
oaire.citationVolume72por
dc.date.updated2024-03-21T11:43:53Z-
dc.identifier.doi10.1109/TC.2022.3218995por
dc.subject.fosEngenharia e Tecnologia::Engenharia Eletrotécnica, Eletrónica e Informáticapor
sdum.export.identifier13599-
sdum.journalIEEE Transactions on Computerspor
oaire.versionAMpor
Aparece nas coleções:HASLab - Artigos em revistas internacionais

Ficheiros deste registo:
Ficheiro Descrição TamanhoFormato 
TC3218995-accepted-version.pdf630,24 kBAdobe PDFVer/Abrir

Partilhe no FacebookPartilhe no TwitterPartilhe no DeliciousPartilhe no LinkedInPartilhe no DiggAdicionar ao Google BookmarksPartilhe no MySpacePartilhe no Orkut
Exporte no formato BibTex mendeley Exporte no formato Endnote Adicione ao seu ORCID