Nonoblivious hashing

Amos Fiat, Moni Naor, Jeanette P. Schmidt, Alan Siegel

Research output: Contribution to journalArticle

Abstract

Nonoblivious hashing, where information gathered from unsuccessful probes is used to modify subsequent probe strategy, is introduced and used to obtain the following results for static lookup on full tables: (1) An O(1)-time worst-case scheme that uses only logarithmic additional memory, (and no memory when the domain size is linear in the table size), which improves upon previously linear space requirements. (2) An almost sure O(1)-time probabilistic worst-case scheme, which uses no additional memory and which improves upon previously logarithmic time requirements. (3) Enhancements to hashing: (1) and (2) are solved for multikey records, where search can be performed under any key in time O(1); these schemes also permit properties, such as nearest neighbor and rank, to be determined in logarithmic time.

Original languageEnglish (US)
Pages (from-to)764-782
Number of pages19
JournalJournal of the ACM
Volume39
Issue number4
DOIs
StatePublished - Oct 1992

Fingerprint

Hashing
Data storage equipment
Logarithmic
Probe
Requirements
Linear Space
Tables
Nearest Neighbor
Table
Enhancement

ASJC Scopus subject areas

  • Computational Theory and Mathematics
  • Computer Graphics and Computer-Aided Design
  • Hardware and Architecture
  • Information Systems
  • Software
  • Theoretical Computer Science

Cite this

Fiat, A., Naor, M., Schmidt, J. P., & Siegel, A. (1992). Nonoblivious hashing. Journal of the ACM, 39(4), 764-782. https://doi.org/10.1145/146585.146591

Nonoblivious hashing. / Fiat, Amos; Naor, Moni; Schmidt, Jeanette P.; Siegel, Alan.

In: Journal of the ACM, Vol. 39, No. 4, 10.1992, p. 764-782.

Research output: Contribution to journalArticle

Fiat, A, Naor, M, Schmidt, JP & Siegel, A 1992, 'Nonoblivious hashing', Journal of the ACM, vol. 39, no. 4, pp. 764-782. https://doi.org/10.1145/146585.146591
Fiat, Amos ; Naor, Moni ; Schmidt, Jeanette P. ; Siegel, Alan. / Nonoblivious hashing. In: Journal of the ACM. 1992 ; Vol. 39, No. 4. pp. 764-782.
@article{21cdb95794b7444ab3f443b0ce4ffd9a,
title = "Nonoblivious hashing",
abstract = "Nonoblivious hashing, where information gathered from unsuccessful probes is used to modify subsequent probe strategy, is introduced and used to obtain the following results for static lookup on full tables: (1) An O(1)-time worst-case scheme that uses only logarithmic additional memory, (and no memory when the domain size is linear in the table size), which improves upon previously linear space requirements. (2) An almost sure O(1)-time probabilistic worst-case scheme, which uses no additional memory and which improves upon previously logarithmic time requirements. (3) Enhancements to hashing: (1) and (2) are solved for multikey records, where search can be performed under any key in time O(1); these schemes also permit properties, such as nearest neighbor and rank, to be determined in logarithmic time.",
author = "Amos Fiat and Moni Naor and Schmidt, {Jeanette P.} and Alan Siegel",
year = "1992",
month = "10",
doi = "10.1145/146585.146591",
language = "English (US)",
volume = "39",
pages = "764--782",
journal = "Journal of the ACM",
issn = "0004-5411",
publisher = "Association for Computing Machinery (ACM)",
number = "4",

}

TY - JOUR

T1 - Nonoblivious hashing

AU - Fiat, Amos

AU - Naor, Moni

AU - Schmidt, Jeanette P.

AU - Siegel, Alan

PY - 1992/10

Y1 - 1992/10

N2 - Nonoblivious hashing, where information gathered from unsuccessful probes is used to modify subsequent probe strategy, is introduced and used to obtain the following results for static lookup on full tables: (1) An O(1)-time worst-case scheme that uses only logarithmic additional memory, (and no memory when the domain size is linear in the table size), which improves upon previously linear space requirements. (2) An almost sure O(1)-time probabilistic worst-case scheme, which uses no additional memory and which improves upon previously logarithmic time requirements. (3) Enhancements to hashing: (1) and (2) are solved for multikey records, where search can be performed under any key in time O(1); these schemes also permit properties, such as nearest neighbor and rank, to be determined in logarithmic time.

AB - Nonoblivious hashing, where information gathered from unsuccessful probes is used to modify subsequent probe strategy, is introduced and used to obtain the following results for static lookup on full tables: (1) An O(1)-time worst-case scheme that uses only logarithmic additional memory, (and no memory when the domain size is linear in the table size), which improves upon previously linear space requirements. (2) An almost sure O(1)-time probabilistic worst-case scheme, which uses no additional memory and which improves upon previously logarithmic time requirements. (3) Enhancements to hashing: (1) and (2) are solved for multikey records, where search can be performed under any key in time O(1); these schemes also permit properties, such as nearest neighbor and rank, to be determined in logarithmic time.

UR - http://www.scopus.com/inward/record.url?scp=0026931776&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0026931776&partnerID=8YFLogxK

U2 - 10.1145/146585.146591

DO - 10.1145/146585.146591

M3 - Article

AN - SCOPUS:0026931776

VL - 39

SP - 764

EP - 782

JO - Journal of the ACM

JF - Journal of the ACM

SN - 0004-5411

IS - 4

ER -