Social network regularized Sparse Linear Model for Top-N recommendation

Xiaodong Feng, Ankit Sharma, Jaideep Srivastava, Sen Wu, Zhiwei Tang

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

Social recommendation techniques have been developed to employ user's social connections for both rating prediction and Top-N recommendation. However, they are mostly using social network enhanced matrix factorization (MF) where the objective is to minimize the prediction error of rating scores, which makes it impractical and unsuccessful for Top-N recommendation. This paper thus focuses on developing more effective methods to utilize social network information for Top-N recommendation. Social network regularized Sparse LInear Model (SocSLIM) with its extensions incorporating local learning (LocSocSLIM) to improve efficiency are proposed. SocSLIM learns sparse coefficient matrix for users by solving a sparse representation problem over user-item rating/purchase matrix and user-user social network's adjacency matrix at the same time by sharing coefficient matrix. The coefficient matrix is used to predict the recommendation scores, which are then combined with a proposed item based Distance regularized Sparse LInear Model (DSLIM) to generate recommendations for the users. The experimental results demonstrate that SocSLIM effectively uses the social information to outperform the state-of-the-art methods by at least 12%. Moreover, the local weight learning extension LocSocSLIM significantly improves the efficiency up to 10 times as compared to SocSLIM as the original SLIM while achieving the close performance guarantees.

Original languageEnglish
Pages (from-to)5-15
Number of pages11
JournalEngineering Applications of Artificial Intelligence
Volume51
DOIs
Publication statusPublished - 1 May 2016
Externally publishedYes

Fingerprint

Factorization

Keywords

  • Local learning
  • Social network
  • Sparse Linear Model
  • Top-N recommendation
  • User modeling

ASJC Scopus subject areas

  • Artificial Intelligence
  • Control and Systems Engineering
  • Electrical and Electronic Engineering

Cite this

Social network regularized Sparse Linear Model for Top-N recommendation. / Feng, Xiaodong; Sharma, Ankit; Srivastava, Jaideep; Wu, Sen; Tang, Zhiwei.

In: Engineering Applications of Artificial Intelligence, Vol. 51, 01.05.2016, p. 5-15.

Research output: Contribution to journalArticle

@article{b15d5286465c4534be7a6e08f4dd7741,
title = "Social network regularized Sparse Linear Model for Top-N recommendation",
abstract = "Social recommendation techniques have been developed to employ user's social connections for both rating prediction and Top-N recommendation. However, they are mostly using social network enhanced matrix factorization (MF) where the objective is to minimize the prediction error of rating scores, which makes it impractical and unsuccessful for Top-N recommendation. This paper thus focuses on developing more effective methods to utilize social network information for Top-N recommendation. Social network regularized Sparse LInear Model (SocSLIM) with its extensions incorporating local learning (LocSocSLIM) to improve efficiency are proposed. SocSLIM learns sparse coefficient matrix for users by solving a sparse representation problem over user-item rating/purchase matrix and user-user social network's adjacency matrix at the same time by sharing coefficient matrix. The coefficient matrix is used to predict the recommendation scores, which are then combined with a proposed item based Distance regularized Sparse LInear Model (DSLIM) to generate recommendations for the users. The experimental results demonstrate that SocSLIM effectively uses the social information to outperform the state-of-the-art methods by at least 12{\%}. Moreover, the local weight learning extension LocSocSLIM significantly improves the efficiency up to 10 times as compared to SocSLIM as the original SLIM while achieving the close performance guarantees.",
keywords = "Local learning, Social network, Sparse Linear Model, Top-N recommendation, User modeling",
author = "Xiaodong Feng and Ankit Sharma and Jaideep Srivastava and Sen Wu and Zhiwei Tang",
year = "2016",
month = "5",
day = "1",
doi = "10.1016/j.engappai.2016.01.019",
language = "English",
volume = "51",
pages = "5--15",
journal = "Engineering Applications of Artificial Intelligence",
issn = "0952-1976",
publisher = "Elsevier Limited",

}

TY - JOUR

T1 - Social network regularized Sparse Linear Model for Top-N recommendation

AU - Feng, Xiaodong

AU - Sharma, Ankit

AU - Srivastava, Jaideep

AU - Wu, Sen

AU - Tang, Zhiwei

PY - 2016/5/1

Y1 - 2016/5/1

N2 - Social recommendation techniques have been developed to employ user's social connections for both rating prediction and Top-N recommendation. However, they are mostly using social network enhanced matrix factorization (MF) where the objective is to minimize the prediction error of rating scores, which makes it impractical and unsuccessful for Top-N recommendation. This paper thus focuses on developing more effective methods to utilize social network information for Top-N recommendation. Social network regularized Sparse LInear Model (SocSLIM) with its extensions incorporating local learning (LocSocSLIM) to improve efficiency are proposed. SocSLIM learns sparse coefficient matrix for users by solving a sparse representation problem over user-item rating/purchase matrix and user-user social network's adjacency matrix at the same time by sharing coefficient matrix. The coefficient matrix is used to predict the recommendation scores, which are then combined with a proposed item based Distance regularized Sparse LInear Model (DSLIM) to generate recommendations for the users. The experimental results demonstrate that SocSLIM effectively uses the social information to outperform the state-of-the-art methods by at least 12%. Moreover, the local weight learning extension LocSocSLIM significantly improves the efficiency up to 10 times as compared to SocSLIM as the original SLIM while achieving the close performance guarantees.

AB - Social recommendation techniques have been developed to employ user's social connections for both rating prediction and Top-N recommendation. However, they are mostly using social network enhanced matrix factorization (MF) where the objective is to minimize the prediction error of rating scores, which makes it impractical and unsuccessful for Top-N recommendation. This paper thus focuses on developing more effective methods to utilize social network information for Top-N recommendation. Social network regularized Sparse LInear Model (SocSLIM) with its extensions incorporating local learning (LocSocSLIM) to improve efficiency are proposed. SocSLIM learns sparse coefficient matrix for users by solving a sparse representation problem over user-item rating/purchase matrix and user-user social network's adjacency matrix at the same time by sharing coefficient matrix. The coefficient matrix is used to predict the recommendation scores, which are then combined with a proposed item based Distance regularized Sparse LInear Model (DSLIM) to generate recommendations for the users. The experimental results demonstrate that SocSLIM effectively uses the social information to outperform the state-of-the-art methods by at least 12%. Moreover, the local weight learning extension LocSocSLIM significantly improves the efficiency up to 10 times as compared to SocSLIM as the original SLIM while achieving the close performance guarantees.

KW - Local learning

KW - Social network

KW - Sparse Linear Model

KW - Top-N recommendation

KW - User modeling

UR - http://www.scopus.com/inward/record.url?scp=84961627445&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84961627445&partnerID=8YFLogxK

U2 - 10.1016/j.engappai.2016.01.019

DO - 10.1016/j.engappai.2016.01.019

M3 - Article

VL - 51

SP - 5

EP - 15

JO - Engineering Applications of Artificial Intelligence

JF - Engineering Applications of Artificial Intelligence

SN - 0952-1976

ER -