Abstract

Network based recommendation systems leverage the topology of the underlying graph and the current user context to rank objects in the database. Random-walk based techniques, such as PageRank, encode the structure of the graph in the form of a transition matrix of a stochastic process from which the significances of the nodes in the graph are inferred. Personalized PageRank (PPR) techniques complement this with a seed node set which serves as the personalization context. In this paper, we note (and experimentally show) that PPR algorithms that do not differentiate among the seed nodes may not properly rank nodes in situations where the seed set is incomplete and/or noisy. To tackle this problem, we propose alternative robust personalized PageRank (RPR) strategies, which are insensitive to noise in the set of seed nodes and in which the rankings are not overly biased towards the seed nodes. In particular, we show that novel teleportation discounting and seed-set maximal PPR techniques help eliminate harmful bias of individual seed nodes and provide effective seed differentiation to lead to more accurate rankings.

Original languageEnglish (US)
Title of host publicationASONAM 2014 - Proceedings of the 2014 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages216-223
Number of pages8
ISBN (Print)9781479958771
DOIs
StatePublished - Oct 10 2014
Event2014 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining, ASONAM 2014 - Beijing, China
Duration: Aug 17 2014Aug 20 2014

Other

Other2014 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining, ASONAM 2014
CountryChina
CityBeijing
Period8/17/148/20/14

Fingerprint

Seed
Recommender systems
Random processes
Topology

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Computer Science Applications

Cite this

Huang, S., Li, X., Candan, K., & Sapino, M. L. (2014). 'Can you really trust that seed?': Reducing the impact of seed noise in personalized PageRank. In ASONAM 2014 - Proceedings of the 2014 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (pp. 216-223). [6921586] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ASONAM.2014.6921586

'Can you really trust that seed?' : Reducing the impact of seed noise in personalized PageRank. / Huang, Shengyu; Li, Xinsheng; Candan, Kasim; Sapino, Maria Luisa.

ASONAM 2014 - Proceedings of the 2014 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining. Institute of Electrical and Electronics Engineers Inc., 2014. p. 216-223 6921586.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Huang, S, Li, X, Candan, K & Sapino, ML 2014, 'Can you really trust that seed?': Reducing the impact of seed noise in personalized PageRank. in ASONAM 2014 - Proceedings of the 2014 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining., 6921586, Institute of Electrical and Electronics Engineers Inc., pp. 216-223, 2014 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining, ASONAM 2014, Beijing, China, 8/17/14. https://doi.org/10.1109/ASONAM.2014.6921586
Huang S, Li X, Candan K, Sapino ML. 'Can you really trust that seed?': Reducing the impact of seed noise in personalized PageRank. In ASONAM 2014 - Proceedings of the 2014 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining. Institute of Electrical and Electronics Engineers Inc. 2014. p. 216-223. 6921586 https://doi.org/10.1109/ASONAM.2014.6921586
Huang, Shengyu ; Li, Xinsheng ; Candan, Kasim ; Sapino, Maria Luisa. / 'Can you really trust that seed?' : Reducing the impact of seed noise in personalized PageRank. ASONAM 2014 - Proceedings of the 2014 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining. Institute of Electrical and Electronics Engineers Inc., 2014. pp. 216-223
@inproceedings{5db33923df8b4540b0c3c16ac1536b99,
title = "'Can you really trust that seed?': Reducing the impact of seed noise in personalized PageRank",
abstract = "Network based recommendation systems leverage the topology of the underlying graph and the current user context to rank objects in the database. Random-walk based techniques, such as PageRank, encode the structure of the graph in the form of a transition matrix of a stochastic process from which the significances of the nodes in the graph are inferred. Personalized PageRank (PPR) techniques complement this with a seed node set which serves as the personalization context. In this paper, we note (and experimentally show) that PPR algorithms that do not differentiate among the seed nodes may not properly rank nodes in situations where the seed set is incomplete and/or noisy. To tackle this problem, we propose alternative robust personalized PageRank (RPR) strategies, which are insensitive to noise in the set of seed nodes and in which the rankings are not overly biased towards the seed nodes. In particular, we show that novel teleportation discounting and seed-set maximal PPR techniques help eliminate harmful bias of individual seed nodes and provide effective seed differentiation to lead to more accurate rankings.",
author = "Shengyu Huang and Xinsheng Li and Kasim Candan and Sapino, {Maria Luisa}",
year = "2014",
month = "10",
day = "10",
doi = "10.1109/ASONAM.2014.6921586",
language = "English (US)",
isbn = "9781479958771",
pages = "216--223",
booktitle = "ASONAM 2014 - Proceedings of the 2014 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - 'Can you really trust that seed?'

T2 - Reducing the impact of seed noise in personalized PageRank

AU - Huang, Shengyu

AU - Li, Xinsheng

AU - Candan, Kasim

AU - Sapino, Maria Luisa

PY - 2014/10/10

Y1 - 2014/10/10

N2 - Network based recommendation systems leverage the topology of the underlying graph and the current user context to rank objects in the database. Random-walk based techniques, such as PageRank, encode the structure of the graph in the form of a transition matrix of a stochastic process from which the significances of the nodes in the graph are inferred. Personalized PageRank (PPR) techniques complement this with a seed node set which serves as the personalization context. In this paper, we note (and experimentally show) that PPR algorithms that do not differentiate among the seed nodes may not properly rank nodes in situations where the seed set is incomplete and/or noisy. To tackle this problem, we propose alternative robust personalized PageRank (RPR) strategies, which are insensitive to noise in the set of seed nodes and in which the rankings are not overly biased towards the seed nodes. In particular, we show that novel teleportation discounting and seed-set maximal PPR techniques help eliminate harmful bias of individual seed nodes and provide effective seed differentiation to lead to more accurate rankings.

AB - Network based recommendation systems leverage the topology of the underlying graph and the current user context to rank objects in the database. Random-walk based techniques, such as PageRank, encode the structure of the graph in the form of a transition matrix of a stochastic process from which the significances of the nodes in the graph are inferred. Personalized PageRank (PPR) techniques complement this with a seed node set which serves as the personalization context. In this paper, we note (and experimentally show) that PPR algorithms that do not differentiate among the seed nodes may not properly rank nodes in situations where the seed set is incomplete and/or noisy. To tackle this problem, we propose alternative robust personalized PageRank (RPR) strategies, which are insensitive to noise in the set of seed nodes and in which the rankings are not overly biased towards the seed nodes. In particular, we show that novel teleportation discounting and seed-set maximal PPR techniques help eliminate harmful bias of individual seed nodes and provide effective seed differentiation to lead to more accurate rankings.

UR - http://www.scopus.com/inward/record.url?scp=84911095295&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84911095295&partnerID=8YFLogxK

U2 - 10.1109/ASONAM.2014.6921586

DO - 10.1109/ASONAM.2014.6921586

M3 - Conference contribution

AN - SCOPUS:84911095295

SN - 9781479958771

SP - 216

EP - 223

BT - ASONAM 2014 - Proceedings of the 2014 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining

PB - Institute of Electrical and Electronics Engineers Inc.

ER -