Sketching algorithms for approximating rank correlations in collaborative filtering systems

Yoram Bachrach; Ralf Herbrich; Ely Porat

doi:10.1007/978-3-642-03784-9_34

Sketching algorithms for approximating rank correlations in collaborative filtering systems

Yoram Bachrach, Ralf Herbrich, Ely Porat

Department of Computer Science

Microsoft USA

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

10 Scopus citations

Abstract

Collaborative filtering (CF) shares information between users to provide each with recommendations. Previous work suggests using sketching techniques to handle massive data sets in CF systems, but only allows testing whether users have a high proportion of items they have both ranked. We show how to determine the correlation between the rankings of two users, using concise "sketches" of the rankings. The sketches allow approximating Kendall's Tau, a known rank correlation, with high accuracy ε and high confidence 1 - δ. The required sketch size is logarithmic in the confidence and polynomial in the accuracy.

Original language	English
Title of host publication	String Processing and Information Retrieval - 16th International Symposium, SPIRE 2009, Proceedings
Pages	344-352
Number of pages	9
DOIs	https://doi.org/10.1007/978-3-642-03784-9_34
State	Published - 2009
Event	16th International Symposium on String Processing and Information Retrieval, SPIRE 2009 - Saariselka, Finland Duration: 25 Aug 2009 → 27 Aug 2009

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	5721 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	16th International Symposium on String Processing and Information Retrieval, SPIRE 2009
Country/Territory	Finland
City	Saariselka
Period	25/08/09 → 27/08/09

Access to Document

10.1007/978-3-642-03784-9_34

Cite this

Bachrach, Y., Herbrich, R., & Porat, E. (2009). Sketching algorithms for approximating rank correlations in collaborative filtering systems. In String Processing and Information Retrieval - 16th International Symposium, SPIRE 2009, Proceedings (pp. 344-352). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5721 LNCS). https://doi.org/10.1007/978-3-642-03784-9_34

Bachrach, Yoram ; Herbrich, Ralf ; Porat, Ely. / Sketching algorithms for approximating rank correlations in collaborative filtering systems. String Processing and Information Retrieval - 16th International Symposium, SPIRE 2009, Proceedings. 2009. pp. 344-352 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{9deaa98a3ff6421aadbeea89d611d8ba,

title = "Sketching algorithms for approximating rank correlations in collaborative filtering systems",

abstract = "Collaborative filtering (CF) shares information between users to provide each with recommendations. Previous work suggests using sketching techniques to handle massive data sets in CF systems, but only allows testing whether users have a high proportion of items they have both ranked. We show how to determine the correlation between the rankings of two users, using concise {"}sketches{"} of the rankings. The sketches allow approximating Kendall's Tau, a known rank correlation, with high accuracy ε and high confidence 1 - δ. The required sketch size is logarithmic in the confidence and polynomial in the accuracy.",

author = "Yoram Bachrach and Ralf Herbrich and Ely Porat",

year = "2009",

doi = "10.1007/978-3-642-03784-9_34",

language = "אנגלית",

isbn = "3642037836",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

pages = "344--352",

booktitle = "String Processing and Information Retrieval - 16th International Symposium, SPIRE 2009, Proceedings",

note = "16th International Symposium on String Processing and Information Retrieval, SPIRE 2009 ; Conference date: 25-08-2009 Through 27-08-2009",

}

Bachrach, Y, Herbrich, R & Porat, E 2009, Sketching algorithms for approximating rank correlations in collaborative filtering systems. in String Processing and Information Retrieval - 16th International Symposium, SPIRE 2009, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 5721 LNCS, pp. 344-352, 16th International Symposium on String Processing and Information Retrieval, SPIRE 2009, Saariselka, Finland, 25/08/09. https://doi.org/10.1007/978-3-642-03784-9_34

Sketching algorithms for approximating rank correlations in collaborative filtering systems. / Bachrach, Yoram; Herbrich, Ralf; Porat, Ely.
String Processing and Information Retrieval - 16th International Symposium, SPIRE 2009, Proceedings. 2009. p. 344-352 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5721 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Sketching algorithms for approximating rank correlations in collaborative filtering systems

AU - Bachrach, Yoram

AU - Herbrich, Ralf

AU - Porat, Ely

PY - 2009

Y1 - 2009

N2 - Collaborative filtering (CF) shares information between users to provide each with recommendations. Previous work suggests using sketching techniques to handle massive data sets in CF systems, but only allows testing whether users have a high proportion of items they have both ranked. We show how to determine the correlation between the rankings of two users, using concise "sketches" of the rankings. The sketches allow approximating Kendall's Tau, a known rank correlation, with high accuracy ε and high confidence 1 - δ. The required sketch size is logarithmic in the confidence and polynomial in the accuracy.

AB - Collaborative filtering (CF) shares information between users to provide each with recommendations. Previous work suggests using sketching techniques to handle massive data sets in CF systems, but only allows testing whether users have a high proportion of items they have both ranked. We show how to determine the correlation between the rankings of two users, using concise "sketches" of the rankings. The sketches allow approximating Kendall's Tau, a known rank correlation, with high accuracy ε and high confidence 1 - δ. The required sketch size is logarithmic in the confidence and polynomial in the accuracy.

UR - http://www.scopus.com/inward/record.url?scp=70350625168&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-03784-9_34

DO - 10.1007/978-3-642-03784-9_34

M3 - ???researchoutput.researchoutputtypes.contributiontobookanthology.conference???

AN - SCOPUS:70350625168

SN - 3642037836

SN - 9783642037832

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 344

EP - 352

BT - String Processing and Information Retrieval - 16th International Symposium, SPIRE 2009, Proceedings

T2 - 16th International Symposium on String Processing and Information Retrieval, SPIRE 2009

Y2 - 25 August 2009 through 27 August 2009

ER -

Bachrach Y, Herbrich R, Porat E. Sketching algorithms for approximating rank correlations in collaborative filtering systems. In String Processing and Information Retrieval - 16th International Symposium, SPIRE 2009, Proceedings. 2009. p. 344-352. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-642-03784-9_34

Sketching algorithms for approximating rank correlations in collaborative filtering systems

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this