RLT: Residual-loop training in collaborative filtering for combining factorization and global-local neighborhood

Lei Li; Weike Pan; Li Chen; Zhong Ming

doi:10.1007/978-3-319-94289-6_21

RLT: Residual-loop training in collaborative filtering for combining factorization and global-local neighborhood

Lei Li, Weike Pan^*, Li Chen, Zhong Ming

^*Corresponding author for this work

Department of Computer Science

Research output: Chapter in book/report/conference proceeding › Conference proceeding › peer-review

1 Citation (Scopus)

Abstract

Collaborative filtering (CF) is an important recommendation problem focusing on predicting users’ future preferences by exploiting their historical tastes. One typical training paradigm for this problem is called residual training (RT), which is usually built on two basic components of factorization- and local neighborhood-based methods in a sequential manner. RT has been well recognized with the ability of achieving higher recommendation accuracy than either factorization- or neighborhood-based method. In this paper, we design a new residual training paradigm called residual-loop training (RLT), which aims to fully exploit the complementarity of factorization, global neighborhood and local neighborhood in one single algorithm. Experimental results on three public datasets show the promising results of our RLT compared with several state-of-the-art methods.

Original language	English
Title of host publication	Web Services – ICWS 2018
Subtitle of host publication	25th International Conference, Held as Part of the Services Conference Federation, SCF 2018, Proceedings
Editors	Hai Jin, Qingyang Wang, Liang-Jie Zhang
Publisher	Springer Cham
Pages	326-336
Number of pages	11
Edition	1st
ISBN (Electronic)	9783319942896
ISBN (Print)	9783319942889
DOIs	https://doi.org/10.1007/978-3-319-94289-6_21
Publication status	Published - 19 Jun 2018
Event	25th International Conference on Web Services, ICWS 2018 Held as Part of the Services Conference Federation, SCF 2018 - Seattle, United States Duration: 25 Jun 2018 → 30 Jun 2018

Publication series

Name	Lecture Notes in Computer Science
Volume	10966
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Name	Programming and Software Engineering

Name	ICWS: International Conference on Web Services

Conference

Conference	25th International Conference on Web Services, ICWS 2018 Held as Part of the Services Conference Federation, SCF 2018
Country/Territory	United States
City	Seattle
Period	25/06/18 → 30/06/18

Scopus Subject Areas

Theoretical Computer Science
General Computer Science

User-Defined Keywords

Collaborative filtering
Residual training
Residual-loop training

Access to Document

10.1007/978-3-319-94289-6_21

Cite this

Li, L., Pan, W., Chen, L., & Ming, Z. (2018). RLT: Residual-loop training in collaborative filtering for combining factorization and global-local neighborhood. In H. Jin, Q. Wang, & L.-J. Zhang (Eds.), Web Services – ICWS 2018: 25th International Conference, Held as Part of the Services Conference Federation, SCF 2018, Proceedings (1st ed., pp. 326-336). (Lecture Notes in Computer Science; Vol. 10966), (Programming and Software Engineering), (ICWS: International Conference on Web Services). Springer Cham. https://doi.org/10.1007/978-3-319-94289-6_21

Li, Lei ; Pan, Weike ; Chen, Li et al. / RLT : Residual-loop training in collaborative filtering for combining factorization and global-local neighborhood. Web Services – ICWS 2018: 25th International Conference, Held as Part of the Services Conference Federation, SCF 2018, Proceedings. editor / Hai Jin ; Qingyang Wang ; Liang-Jie Zhang. 1st. ed. Springer Cham, 2018. pp. 326-336 (Lecture Notes in Computer Science). (Programming and Software Engineering). (ICWS: International Conference on Web Services).

@inproceedings{6c1b25062c054239b6a88b2c2853c81d,

title = "RLT: Residual-loop training in collaborative filtering for combining factorization and global-local neighborhood",

abstract = "Collaborative filtering (CF) is an important recommendation problem focusing on predicting users{\textquoteright} future preferences by exploiting their historical tastes. One typical training paradigm for this problem is called residual training (RT), which is usually built on two basic components of factorization- and local neighborhood-based methods in a sequential manner. RT has been well recognized with the ability of achieving higher recommendation accuracy than either factorization- or neighborhood-based method. In this paper, we design a new residual training paradigm called residual-loop training (RLT), which aims to fully exploit the complementarity of factorization, global neighborhood and local neighborhood in one single algorithm. Experimental results on three public datasets show the promising results of our RLT compared with several state-of-the-art methods.",

keywords = "Collaborative filtering, Residual training, Residual-loop training",

author = "Lei Li and Weike Pan and Li Chen and Zhong Ming",

note = "Funding Information: Acknowledgement. We thank the support of National Natural Science Foundation of China Nos. 61502307 and 61672358, Hong Kong RGC under the project RGC/HKBU12200415, and Natural Science Foundation of Guangdong Province No. 2016A030313038. Weike Pan and Zhong Ming are the corresponding authors for this work.; 25th International Conference on Web Services, ICWS 2018 Held as Part of the Services Conference Federation, SCF 2018 ; Conference date: 25-06-2018 Through 30-06-2018",

year = "2018",

month = jun,

day = "19",

doi = "10.1007/978-3-319-94289-6_21",

language = "English",

isbn = "9783319942889",

series = "Lecture Notes in Computer Science",

publisher = "Springer Cham",

pages = "326--336",

editor = "Hai Jin and Qingyang Wang and Liang-Jie Zhang",

booktitle = "Web Services – ICWS 2018",

edition = "1st",

}

Li, L, Pan, W, Chen, L & Ming, Z 2018, RLT: Residual-loop training in collaborative filtering for combining factorization and global-local neighborhood. in H Jin, Q Wang & L-J Zhang (eds), Web Services – ICWS 2018: 25th International Conference, Held as Part of the Services Conference Federation, SCF 2018, Proceedings. 1st edn, Lecture Notes in Computer Science, vol. 10966, Programming and Software Engineering, ICWS: International Conference on Web Services, Springer Cham, pp. 326-336, 25th International Conference on Web Services, ICWS 2018 Held as Part of the Services Conference Federation, SCF 2018, Seattle, United States, 25/06/18. https://doi.org/10.1007/978-3-319-94289-6_21

RLT: Residual-loop training in collaborative filtering for combining factorization and global-local neighborhood. / Li, Lei; Pan, Weike; Chen, Li et al.
Web Services – ICWS 2018: 25th International Conference, Held as Part of the Services Conference Federation, SCF 2018, Proceedings. ed. / Hai Jin; Qingyang Wang; Liang-Jie Zhang. 1st. ed. Springer Cham, 2018. p. 326-336 (Lecture Notes in Computer Science; Vol. 10966), (Programming and Software Engineering), (ICWS: International Conference on Web Services).

Research output: Chapter in book/report/conference proceeding › Conference proceeding › peer-review

TY - GEN

T1 - RLT

T2 - 25th International Conference on Web Services, ICWS 2018 Held as Part of the Services Conference Federation, SCF 2018

AU - Li, Lei

AU - Pan, Weike

AU - Chen, Li

AU - Ming, Zhong

N1 - Funding Information: Acknowledgement. We thank the support of National Natural Science Foundation of China Nos. 61502307 and 61672358, Hong Kong RGC under the project RGC/HKBU12200415, and Natural Science Foundation of Guangdong Province No. 2016A030313038. Weike Pan and Zhong Ming are the corresponding authors for this work.

PY - 2018/6/19

Y1 - 2018/6/19

N2 - Collaborative filtering (CF) is an important recommendation problem focusing on predicting users’ future preferences by exploiting their historical tastes. One typical training paradigm for this problem is called residual training (RT), which is usually built on two basic components of factorization- and local neighborhood-based methods in a sequential manner. RT has been well recognized with the ability of achieving higher recommendation accuracy than either factorization- or neighborhood-based method. In this paper, we design a new residual training paradigm called residual-loop training (RLT), which aims to fully exploit the complementarity of factorization, global neighborhood and local neighborhood in one single algorithm. Experimental results on three public datasets show the promising results of our RLT compared with several state-of-the-art methods.

AB - Collaborative filtering (CF) is an important recommendation problem focusing on predicting users’ future preferences by exploiting their historical tastes. One typical training paradigm for this problem is called residual training (RT), which is usually built on two basic components of factorization- and local neighborhood-based methods in a sequential manner. RT has been well recognized with the ability of achieving higher recommendation accuracy than either factorization- or neighborhood-based method. In this paper, we design a new residual training paradigm called residual-loop training (RLT), which aims to fully exploit the complementarity of factorization, global neighborhood and local neighborhood in one single algorithm. Experimental results on three public datasets show the promising results of our RLT compared with several state-of-the-art methods.

KW - Collaborative filtering

KW - Residual training

KW - Residual-loop training

UR - http://www.scopus.com/inward/record.url?scp=85049378211&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-94289-6_21

DO - 10.1007/978-3-319-94289-6_21

M3 - Conference proceeding

AN - SCOPUS:85049378211

SN - 9783319942889

T3 - Lecture Notes in Computer Science

SP - 326

EP - 336

BT - Web Services – ICWS 2018

A2 - Jin, Hai

A2 - Wang, Qingyang

A2 - Zhang, Liang-Jie

PB - Springer Cham

Y2 - 25 June 2018 through 30 June 2018

ER -

Li L, Pan W, Chen L, Ming Z. RLT: Residual-loop training in collaborative filtering for combining factorization and global-local neighborhood. In Jin H, Wang Q, Zhang LJ, editors, Web Services – ICWS 2018: 25th International Conference, Held as Part of the Services Conference Federation, SCF 2018, Proceedings. 1st ed. Springer Cham. 2018. p. 326-336. (Lecture Notes in Computer Science). (Programming and Software Engineering). (ICWS: International Conference on Web Services). doi: 10.1007/978-3-319-94289-6_21

RLT: Residual-loop training in collaborative filtering for combining factorization and global-local neighborhood

Abstract

Publication series

Conference

Scopus Subject Areas

User-Defined Keywords

Access to Document

Other files and links

Fingerprint

Cite this