Collaborative filtering using random neighbours in peer-to-peer networks

Arno Bakker; Elth Ogston; Maarten van Steen

doi:10.1145/1651274.1651288

Collaborative filtering using random neighbours in peer-to-peer networks

Arno Bakker^*, Elth Ogston, Maarten van Steen

^*Corresponding author for this work

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

19 Citations (Scopus)

Abstract

Traditionally, collaborative filtering (CF) algorithms used for recommendation operate on complete knowledge. This makes these algorithms hard to employ in a decentralized context where not all users' ratings can be available at all locations. In this paper we investigate how the well-known neighbourhood-based CF algorithm by Herlocker et al. operates on partial knowledge; that is, how many similar users does the algorithm actually need to produce good recommendations for a given user, and how similar must those users be. We show for the popular MovieLens 1,000,000 and Jester datasets that sufficiently good recommendations can be made based on the ratings of a neighbourhood consisting of a relatively small number of randomly selected users.

Original language	English
Title of host publication	CNIKM '09
Subtitle of host publication	Proceedings of the 1st ACM International Workshop on Complex Networks in Information and Knowledge Management (CNIKM 2009)
Place of Publication	New York, NY
Publisher	ACM Publishing
Pages	67-74
Number of pages	8
ISBN (Print)	978-1-60558-807-0
DOIs	https://doi.org/10.1145/1651274.1651288
Publication status	Published - 1 Dec 2009
Externally published	Yes
Event	1st ACM International Workshop on Complex Networks in Information and Knowledge Management, , CNIKM'09 - Hong Kong, China Duration: 6 Nov 2009 → 6 Nov 2009

Conference

Conference	1st ACM International Workshop on Complex Networks in Information and Knowledge Management, , CNIKM'09
Country/Territory	China
City	Hong Kong
Period	6/11/09 → 6/11/09
Other	Co-located with the 18th ACM International Conference on Information and Knowledge Management, CIKM 2009

Keywords

Collaborative filtering
Metrics
Peer-to-peer networking
Recommender systems

Access to Document

10.1145/1651274.1651288

Cite this

@inproceedings{fa53680c88b44c7881bb8610359acb36,

title = "Collaborative filtering using random neighbours in peer-to-peer networks",

abstract = "Traditionally, collaborative filtering (CF) algorithms used for recommendation operate on complete knowledge. This makes these algorithms hard to employ in a decentralized context where not all users' ratings can be available at all locations. In this paper we investigate how the well-known neighbourhood-based CF algorithm by Herlocker et al. operates on partial knowledge; that is, how many similar users does the algorithm actually need to produce good recommendations for a given user, and how similar must those users be. We show for the popular MovieLens 1,000,000 and Jester datasets that sufficiently good recommendations can be made based on the ratings of a neighbourhood consisting of a relatively small number of randomly selected users.",

keywords = "Collaborative filtering, Metrics, Peer-to-peer networking, Recommender systems",

author = "Arno Bakker and Elth Ogston and {van Steen}, Maarten",

year = "2009",

month = dec,

day = "1",

doi = "10.1145/1651274.1651288",

language = "English",

isbn = "978-1-60558-807-0",

pages = "67--74",

booktitle = "CNIKM '09",

publisher = "ACM Publishing",

note = "1st ACM International Workshop on Complex Networks in Information and Knowledge Management, , CNIKM'09 ; Conference date: 06-11-2009 Through 06-11-2009",

}

Bakker, A, Ogston, E & van Steen, M 2009, Collaborative filtering using random neighbours in peer-to-peer networks. in CNIKM '09 : Proceedings of the 1st ACM International Workshop on Complex Networks in Information and Knowledge Management (CNIKM 2009). ACM Publishing, New York, NY, pp. 67-74, 1st ACM International Workshop on Complex Networks in Information and Knowledge Management, , CNIKM'09, Hong Kong, China, 6/11/09. https://doi.org/10.1145/1651274.1651288

Collaborative filtering using random neighbours in peer-to-peer networks. / Bakker, Arno; Ogston, Elth; van Steen, Maarten.

CNIKM '09 : Proceedings of the 1st ACM International Workshop on Complex Networks in Information and Knowledge Management (CNIKM 2009). New York, NY : ACM Publishing, 2009. p. 67-74.

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

TY - GEN

T1 - Collaborative filtering using random neighbours in peer-to-peer networks

AU - Bakker, Arno

AU - Ogston, Elth

AU - van Steen, Maarten

PY - 2009/12/1

Y1 - 2009/12/1

N2 - Traditionally, collaborative filtering (CF) algorithms used for recommendation operate on complete knowledge. This makes these algorithms hard to employ in a decentralized context where not all users' ratings can be available at all locations. In this paper we investigate how the well-known neighbourhood-based CF algorithm by Herlocker et al. operates on partial knowledge; that is, how many similar users does the algorithm actually need to produce good recommendations for a given user, and how similar must those users be. We show for the popular MovieLens 1,000,000 and Jester datasets that sufficiently good recommendations can be made based on the ratings of a neighbourhood consisting of a relatively small number of randomly selected users.

AB - Traditionally, collaborative filtering (CF) algorithms used for recommendation operate on complete knowledge. This makes these algorithms hard to employ in a decentralized context where not all users' ratings can be available at all locations. In this paper we investigate how the well-known neighbourhood-based CF algorithm by Herlocker et al. operates on partial knowledge; that is, how many similar users does the algorithm actually need to produce good recommendations for a given user, and how similar must those users be. We show for the popular MovieLens 1,000,000 and Jester datasets that sufficiently good recommendations can be made based on the ratings of a neighbourhood consisting of a relatively small number of randomly selected users.

KW - Collaborative filtering

KW - Metrics

KW - Peer-to-peer networking

KW - Recommender systems

UR - http://www.scopus.com/inward/record.url?scp=74049131622&partnerID=8YFLogxK

U2 - 10.1145/1651274.1651288

DO - 10.1145/1651274.1651288

M3 - Conference contribution

AN - SCOPUS:74049131622

SN - 978-1-60558-807-0

SP - 67

EP - 74

BT - CNIKM '09

PB - ACM Publishing

CY - New York, NY

T2 - 1st ACM International Workshop on Complex Networks in Information and Knowledge Management, , CNIKM'09

Y2 - 6 November 2009 through 6 November 2009

ER -

Collaborative filtering using random neighbours in peer-to-peer networks

Abstract

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this