Collaborative Filtering-Based Recommendation of Online Social Voting

Xiwang Yang, Chao Liang, Miao Zhao, Hongwei Wang, Hao Ding, Yong Liu, Yang Li, Junlin Zhang

Research output: Contribution to journalArticle

Abstract

Social voting is an emerging new feature in online social networks. It poses unique challenges and opportunities for recommendation. In this paper, we develop a set of matrix-factorization (MF) and nearest-neighbor (NN)-based recommender systems (RSs) that explore user social network and group affiliation information for social voting recommendation. Through experiments with real social voting traces, we demonstrate that social network and group affiliation information can significantly improve the accuracy of popularity-based voting recommendation, and social network information dominates group affiliation information in NN-based approaches. We also observe that social and group information is much more valuable to cold users than to heavy users. In our experiments, simple metapath-based NN models outperform computation-intensive MF models in hot-voting recommendation, while users' interests for nonhot votings can be better mined by MF models. We further propose a hybrid RS, bagging different single approaches to achieve the best top-k hit rate.

Original languageEnglish (US)
Article number7866820
Pages (from-to)1-13
Number of pages13
JournalIEEE Transactions on Computational Social Systems
Volume4
Issue number1
DOIs
StatePublished - Mar 1 2017

Fingerprint

Collaborative filtering
Collaborative Filtering
Voting
Factorization
voting
Recommendations
Recommender systems
Social Networks
social network
Nearest Neighbor
Matrix Factorization
Recommender Systems
Experiments
Factorization of Matrices
Bagging
experiment
Hits
Hybrid Systems
popularity
Experiment

Keywords

  • Collaborative filtering
  • online social networks (OSNs)
  • recommender systems (RSs)
  • social voting

ASJC Scopus subject areas

  • Modeling and Simulation
  • Social Sciences (miscellaneous)
  • Human-Computer Interaction

Cite this

Collaborative Filtering-Based Recommendation of Online Social Voting. / Yang, Xiwang; Liang, Chao; Zhao, Miao; Wang, Hongwei; Ding, Hao; Liu, Yong; Li, Yang; Zhang, Junlin.

In: IEEE Transactions on Computational Social Systems, Vol. 4, No. 1, 7866820, 01.03.2017, p. 1-13.

Research output: Contribution to journalArticle

Yang, Xiwang ; Liang, Chao ; Zhao, Miao ; Wang, Hongwei ; Ding, Hao ; Liu, Yong ; Li, Yang ; Zhang, Junlin. / Collaborative Filtering-Based Recommendation of Online Social Voting. In: IEEE Transactions on Computational Social Systems. 2017 ; Vol. 4, No. 1. pp. 1-13.
@article{cccdd0bb1e6e4aabab45e107277137e1,
title = "Collaborative Filtering-Based Recommendation of Online Social Voting",
abstract = "Social voting is an emerging new feature in online social networks. It poses unique challenges and opportunities for recommendation. In this paper, we develop a set of matrix-factorization (MF) and nearest-neighbor (NN)-based recommender systems (RSs) that explore user social network and group affiliation information for social voting recommendation. Through experiments with real social voting traces, we demonstrate that social network and group affiliation information can significantly improve the accuracy of popularity-based voting recommendation, and social network information dominates group affiliation information in NN-based approaches. We also observe that social and group information is much more valuable to cold users than to heavy users. In our experiments, simple metapath-based NN models outperform computation-intensive MF models in hot-voting recommendation, while users' interests for nonhot votings can be better mined by MF models. We further propose a hybrid RS, bagging different single approaches to achieve the best top-k hit rate.",
keywords = "Collaborative filtering, online social networks (OSNs), recommender systems (RSs), social voting",
author = "Xiwang Yang and Chao Liang and Miao Zhao and Hongwei Wang and Hao Ding and Yong Liu and Yang Li and Junlin Zhang",
year = "2017",
month = "3",
day = "1",
doi = "10.1109/TCSS.2017.2665122",
language = "English (US)",
volume = "4",
pages = "1--13",
journal = "IEEE Transactions on Computational Social Systems",
issn = "2329-924X",
publisher = "IEEE Systems, Man, and Cybernetics Society",
number = "1",

}

TY - JOUR

T1 - Collaborative Filtering-Based Recommendation of Online Social Voting

AU - Yang, Xiwang

AU - Liang, Chao

AU - Zhao, Miao

AU - Wang, Hongwei

AU - Ding, Hao

AU - Liu, Yong

AU - Li, Yang

AU - Zhang, Junlin

PY - 2017/3/1

Y1 - 2017/3/1

N2 - Social voting is an emerging new feature in online social networks. It poses unique challenges and opportunities for recommendation. In this paper, we develop a set of matrix-factorization (MF) and nearest-neighbor (NN)-based recommender systems (RSs) that explore user social network and group affiliation information for social voting recommendation. Through experiments with real social voting traces, we demonstrate that social network and group affiliation information can significantly improve the accuracy of popularity-based voting recommendation, and social network information dominates group affiliation information in NN-based approaches. We also observe that social and group information is much more valuable to cold users than to heavy users. In our experiments, simple metapath-based NN models outperform computation-intensive MF models in hot-voting recommendation, while users' interests for nonhot votings can be better mined by MF models. We further propose a hybrid RS, bagging different single approaches to achieve the best top-k hit rate.

AB - Social voting is an emerging new feature in online social networks. It poses unique challenges and opportunities for recommendation. In this paper, we develop a set of matrix-factorization (MF) and nearest-neighbor (NN)-based recommender systems (RSs) that explore user social network and group affiliation information for social voting recommendation. Through experiments with real social voting traces, we demonstrate that social network and group affiliation information can significantly improve the accuracy of popularity-based voting recommendation, and social network information dominates group affiliation information in NN-based approaches. We also observe that social and group information is much more valuable to cold users than to heavy users. In our experiments, simple metapath-based NN models outperform computation-intensive MF models in hot-voting recommendation, while users' interests for nonhot votings can be better mined by MF models. We further propose a hybrid RS, bagging different single approaches to achieve the best top-k hit rate.

KW - Collaborative filtering

KW - online social networks (OSNs)

KW - recommender systems (RSs)

KW - social voting

UR - http://www.scopus.com/inward/record.url?scp=85015269237&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85015269237&partnerID=8YFLogxK

U2 - 10.1109/TCSS.2017.2665122

DO - 10.1109/TCSS.2017.2665122

M3 - Article

AN - SCOPUS:85015269237

VL - 4

SP - 1

EP - 13

JO - IEEE Transactions on Computational Social Systems

JF - IEEE Transactions on Computational Social Systems

SN - 2329-924X

IS - 1

M1 - 7866820

ER -