Popularity tendency analysis of ranking-oriented collaborative filtering from the perspective of loss function

Xudong MAO; Qing LI; Haoran XIE; Yanghui RAO

doi:10.1007/978-3-319-05810-8_30

Popularity tendency analysis of ranking-oriented collaborative filtering from the perspective of loss function

Xudong MAO, Qing LI, Haoran XIE, Yanghui RAO

Research output: Book Chapters | Papers in Conference Proceedings › Conference paper (refereed) › Research › peer-review

5 Citations (Scopus)

Abstract

Collaborative filtering (CF) has been the most popular approach for recommender systems in recent years. In order to analyze the property of a ranking-oriented CF algorithm directly and be able to improve its performance, this paper investigates the ranking-oriented CF from the perspective of loss function. To gain the insight into the popular bias problem, we also study the tendency of a CF algorithm in recommending the most popular items, and show that such popularity tendency can be adjusted through setting different parameters in our models. After analyzing two state-of-the-art algorithms, we propose in this paper two models using the generalized logistic loss function and the hinge loss function, respectively. The experimental results show that the proposed methods outperform the state-of-the-art algorithms on two real data sets.

Original language	English
Title of host publication	Database Systems for Advanced Applications
Editors	Sourav S. BHOWMICK, Curtis E. DYRESON, Christian S. JENSEN, Mong Li LEE, Agus MULIANTARA, Bernhard THALHEIM
Place of Publication	Switzerland
Publisher	Springer, Cham
Pages	451-465
Number of pages	15
Volume	8421
ISBN (Electronic)	9783319058108
ISBN (Print)	9783319058092
DOIs	https://doi.org/10.1007/978-3-319-05810-8_30
Publication status	Published - 2014
Externally published	Yes
Event	The 19th International Conference on Database Systems for Advanced Applications - Aerowisata Sanur Beach Hotel, Bali, Indonesia Duration: 21 Apr 2014 → 24 Apr 2014 https://www.comp.nus.edu.sg/~dasfaa14/

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Publisher	Springer-Verlag GmbH and Co. KG
ISSN (Print)	0302-9743

Conference

Conference	The 19th International Conference on Database Systems for Advanced Applications
Abbreviated title	DASFAA 2014
Country/Territory	Indonesia
City	Bali
Period	21/04/14 → 24/04/14
Internet address	https://www.comp.nus.edu.sg/~dasfaa14/

Keywords

Collaborative filtering
loss function
matrix factorization

Access to Document

10.1007/978-3-319-05810-8_30

Cite this

MAO, X., LI, Q., XIE, H., & RAO, Y. (2014). Popularity tendency analysis of ranking-oriented collaborative filtering from the perspective of loss function. In S. S. BHOWMICK, C. E. DYRESON, C. S. JENSEN, M. L. LEE, A. MULIANTARA, & B. THALHEIM (Eds.), Database Systems for Advanced Applications (Vol. 8421, pp. 451-465). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). Springer, Cham. https://doi.org/10.1007/978-3-319-05810-8_30

MAO, Xudong ; LI, Qing ; XIE, Haoran et al. / Popularity tendency analysis of ranking-oriented collaborative filtering from the perspective of loss function. Database Systems for Advanced Applications. editor / Sourav S. BHOWMICK ; Curtis E. DYRESON ; Christian S. JENSEN ; Mong Li LEE ; Agus MULIANTARA ; Bernhard THALHEIM. Vol. 8421 Switzerland : Springer, Cham, 2014. pp. 451-465 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{c450652996784f59b97cb61953a3a768,

title = "Popularity tendency analysis of ranking-oriented collaborative filtering from the perspective of loss function",

abstract = "Collaborative filtering (CF) has been the most popular approach for recommender systems in recent years. In order to analyze the property of a ranking-oriented CF algorithm directly and be able to improve its performance, this paper investigates the ranking-oriented CF from the perspective of loss function. To gain the insight into the popular bias problem, we also study the tendency of a CF algorithm in recommending the most popular items, and show that such popularity tendency can be adjusted through setting different parameters in our models. After analyzing two state-of-the-art algorithms, we propose in this paper two models using the generalized logistic loss function and the hinge loss function, respectively. The experimental results show that the proposed methods outperform the state-of-the-art algorithms on two real data sets.",

keywords = "Collaborative filtering, loss function, matrix factorization",

author = "Xudong MAO and Qing LI and Haoran XIE and Yanghui RAO",

year = "2014",

doi = "10.1007/978-3-319-05810-8_30",

language = "English",

isbn = "9783319058092",

volume = "8421",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer, Cham",

pages = "451--465",

editor = "BHOWMICK, {Sourav S.} and DYRESON, {Curtis E.} and JENSEN, {Christian S.} and LEE, {Mong Li} and Agus MULIANTARA and Bernhard THALHEIM",

booktitle = "Database Systems for Advanced Applications",

note = "The 19th International Conference on Database Systems for Advanced Applications, DASFAA 2014 ; Conference date: 21-04-2014 Through 24-04-2014",

url = "https://www.comp.nus.edu.sg/~dasfaa14/",

}

MAO, X, LI, Q, XIE, H & RAO, Y 2014, Popularity tendency analysis of ranking-oriented collaborative filtering from the perspective of loss function. in SS BHOWMICK, CE DYRESON, CS JENSEN, ML LEE, A MULIANTARA & B THALHEIM (eds), Database Systems for Advanced Applications. vol. 8421, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Springer, Cham, Switzerland, pp. 451-465, The 19th International Conference on Database Systems for Advanced Applications, Bali, Indonesia, 21/04/14. https://doi.org/10.1007/978-3-319-05810-8_30

Popularity tendency analysis of ranking-oriented collaborative filtering from the perspective of loss function. / MAO, Xudong; LI, Qing; XIE, Haoran et al.
Database Systems for Advanced Applications. ed. / Sourav S. BHOWMICK; Curtis E. DYRESON; Christian S. JENSEN; Mong Li LEE; Agus MULIANTARA; Bernhard THALHEIM. Vol. 8421 Switzerland: Springer, Cham, 2014. p. 451-465 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

Research output: Book Chapters | Papers in Conference Proceedings › Conference paper (refereed) › Research › peer-review

TY - GEN

T1 - Popularity tendency analysis of ranking-oriented collaborative filtering from the perspective of loss function

AU - MAO, Xudong

AU - LI, Qing

AU - XIE, Haoran

AU - RAO, Yanghui

PY - 2014

Y1 - 2014

N2 - Collaborative filtering (CF) has been the most popular approach for recommender systems in recent years. In order to analyze the property of a ranking-oriented CF algorithm directly and be able to improve its performance, this paper investigates the ranking-oriented CF from the perspective of loss function. To gain the insight into the popular bias problem, we also study the tendency of a CF algorithm in recommending the most popular items, and show that such popularity tendency can be adjusted through setting different parameters in our models. After analyzing two state-of-the-art algorithms, we propose in this paper two models using the generalized logistic loss function and the hinge loss function, respectively. The experimental results show that the proposed methods outperform the state-of-the-art algorithms on two real data sets.

AB - Collaborative filtering (CF) has been the most popular approach for recommender systems in recent years. In order to analyze the property of a ranking-oriented CF algorithm directly and be able to improve its performance, this paper investigates the ranking-oriented CF from the perspective of loss function. To gain the insight into the popular bias problem, we also study the tendency of a CF algorithm in recommending the most popular items, and show that such popularity tendency can be adjusted through setting different parameters in our models. After analyzing two state-of-the-art algorithms, we propose in this paper two models using the generalized logistic loss function and the hinge loss function, respectively. The experimental results show that the proposed methods outperform the state-of-the-art algorithms on two real data sets.

KW - Collaborative filtering

KW - loss function

KW - matrix factorization

UR - http://www.scopus.com/inward/record.url?scp=84900300094&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-05810-8_30

DO - 10.1007/978-3-319-05810-8_30

M3 - Conference paper (refereed)

AN - SCOPUS:84900300094

SN - 9783319058092

VL - 8421

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 451

EP - 465

BT - Database Systems for Advanced Applications

A2 - BHOWMICK, Sourav S.

A2 - DYRESON, Curtis E.

A2 - JENSEN, Christian S.

A2 - LEE, Mong Li

A2 - MULIANTARA, Agus

A2 - THALHEIM, Bernhard

PB - Springer, Cham

CY - Switzerland

T2 - The 19th International Conference on Database Systems for Advanced Applications

Y2 - 21 April 2014 through 24 April 2014

ER -

MAO X, LI Q, XIE H, RAO Y. Popularity tendency analysis of ranking-oriented collaborative filtering from the perspective of loss function. In BHOWMICK SS, DYRESON CE, JENSEN CS, LEE ML, MULIANTARA A, THALHEIM B, editors, Database Systems for Advanced Applications. Vol. 8421. Switzerland: Springer, Cham. 2014. p. 451-465. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-05810-8_30

Popularity tendency analysis of ranking-oriented collaborative filtering from the perspective of loss function

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this