Improved collaborative filtering

Aviv Nisgav; Boaz Patt-Shamir

doi:10.1007/978-3-642-25591-5_44

Improved collaborative filtering

Aviv Nisgav, Boaz Patt-Shamir

School of Electrical Engineering

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

Abstract

We consider the interactive model of collaborative filtering, where each member of a given set of users has a grade for each object in a given set of objects. The users do not know the grades at start, but a user can probe any object, thereby learning her grade for that object directly. We describe reconstruction algorithms which generate good estimates of all user grades ("preference vectors") using only few probes. To this end, the outcomes of probes are posted on some public "billboard", allowing users to adopt results of probes executed by others. We give two new algorithms for this task under very general assumptions on user preferences: both improve the best known query complexity for reconstruction, and one improving resilience in the presence of many users with esoteric taste.

Original language	English
Title of host publication	Algorithms and Computation - 22nd International Symposium, ISAAC 2011, Proceedings
Pages	425-434
Number of pages	10
DOIs	https://doi.org/10.1007/978-3-642-25591-5_44
State	Published - 2011
Event	22nd International Symposium on Algorithms and Computation, ISAAC 2011 - Yokohama, Japan Duration: 5 Dec 2011 → 8 Dec 2011

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	7074 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	22nd International Symposium on Algorithms and Computation, ISAAC 2011
Country/Territory	Japan
City	Yokohama
Period	5/12/11 → 8/12/11

Access to Document

10.1007/978-3-642-25591-5_44

Cite this

@inproceedings{ebf6a4d9092146efa85fa62538756861,

title = "Improved collaborative filtering",

abstract = "We consider the interactive model of collaborative filtering, where each member of a given set of users has a grade for each object in a given set of objects. The users do not know the grades at start, but a user can probe any object, thereby learning her grade for that object directly. We describe reconstruction algorithms which generate good estimates of all user grades ({"}preference vectors{"}) using only few probes. To this end, the outcomes of probes are posted on some public {"}billboard{"}, allowing users to adopt results of probes executed by others. We give two new algorithms for this task under very general assumptions on user preferences: both improve the best known query complexity for reconstruction, and one improving resilience in the presence of many users with esoteric taste.",

author = "Aviv Nisgav and Boaz Patt-Shamir",

year = "2011",

doi = "10.1007/978-3-642-25591-5_44",

language = "אנגלית",

isbn = "9783642255908",

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

pages = "425--434",

booktitle = "Algorithms and Computation - 22nd International Symposium, ISAAC 2011, Proceedings",

note = "null ; Conference date: 05-12-2011 Through 08-12-2011",

}

Nisgav, A & Patt-Shamir, B 2011, Improved collaborative filtering. in Algorithms and Computation - 22nd International Symposium, ISAAC 2011, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 7074 LNCS, pp. 425-434, 22nd International Symposium on Algorithms and Computation, ISAAC 2011, Yokohama, Japan, 5/12/11. https://doi.org/10.1007/978-3-642-25591-5_44

Improved collaborative filtering. / Nisgav, Aviv; Patt-Shamir, Boaz.

Algorithms and Computation - 22nd International Symposium, ISAAC 2011, Proceedings. 2011. p. 425-434 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7074 LNCS).

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

TY - GEN

T1 - Improved collaborative filtering

AU - Nisgav, Aviv

AU - Patt-Shamir, Boaz

PY - 2011

Y1 - 2011

N2 - We consider the interactive model of collaborative filtering, where each member of a given set of users has a grade for each object in a given set of objects. The users do not know the grades at start, but a user can probe any object, thereby learning her grade for that object directly. We describe reconstruction algorithms which generate good estimates of all user grades ("preference vectors") using only few probes. To this end, the outcomes of probes are posted on some public "billboard", allowing users to adopt results of probes executed by others. We give two new algorithms for this task under very general assumptions on user preferences: both improve the best known query complexity for reconstruction, and one improving resilience in the presence of many users with esoteric taste.

AB - We consider the interactive model of collaborative filtering, where each member of a given set of users has a grade for each object in a given set of objects. The users do not know the grades at start, but a user can probe any object, thereby learning her grade for that object directly. We describe reconstruction algorithms which generate good estimates of all user grades ("preference vectors") using only few probes. To this end, the outcomes of probes are posted on some public "billboard", allowing users to adopt results of probes executed by others. We give two new algorithms for this task under very general assumptions on user preferences: both improve the best known query complexity for reconstruction, and one improving resilience in the presence of many users with esoteric taste.

UR - http://www.scopus.com/inward/record.url?scp=84055191088&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-25591-5_44

DO - 10.1007/978-3-642-25591-5_44

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

AN - SCOPUS:84055191088

SN - 9783642255908

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

SP - 425

EP - 434

BT - Algorithms and Computation - 22nd International Symposium, ISAAC 2011, Proceedings

Y2 - 5 December 2011 through 8 December 2011

ER -

Improved collaborative filtering

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this