Matrix Factorization for Package Recommendations

Agung Toto Wibowo, Advaith Siddharthan, Chenghua Lin, Judith Masthoff

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

3 Citations (Scopus)
6 Downloads (Pure)

Abstract

Research in recommendation systems has to date focused on recommending individual items to users. However there are contexts in which combinations of items need to be recommended, and there has been less research to date on how collaborative methods such as matrix factorization can be applied to such tasks. The research contributions of this paper are threefold. First, we formalize the collaborative package recommendation task as an extension of the standard collaborative recommendation task. Second, we describe and make available a novel package recommendation dataset in the clothes domain, where a combination of a "top" (e.g. a shirt, t-shirt or top) and "bottom'" (e.g. trousers, shorts or skirts) needs to be recommended. Finally, we describe several extensions of matrix factorization to predict user ratings on packages, and report RMSE improvements over the standard matrix factorization approach for recommending combinations of tops and bottoms.
Original languageEnglish
Title of host publicationProceedings of the RecSys 2017 Workshop on Recommendation in Complex Scenarios (ComplexRec 2017)
EditorsToine Bogers, Marijn Koolen, Bamshad Mobasher, Alan Said, Alexander Tuzhilin
PublisherCEUR-WS
Pages23-28
Number of pages5
Volume1892
Publication statusPublished - 18 Aug 2017
EventComplexRec 2017 Workshop: Workshop on Recommendation in Complex Scenarios - Como, Italy
Duration: 31 Aug 201731 Aug 2017

Workshop

WorkshopComplexRec 2017 Workshop
CountryItaly
CityComo
Period31/08/1731/08/17

Fingerprint

Factorization
Recommender systems

Keywords

  • Package Recommendation
  • Matrix Factorization
  • Clothes Domain
  • Collaborative Filtering

Cite this

Wibowo, A. T., Siddharthan, A., Lin, C., & Masthoff, J. (2017). Matrix Factorization for Package Recommendations. In T. Bogers, M. Koolen, B. Mobasher, A. Said, & A. Tuzhilin (Eds.), Proceedings of the RecSys 2017 Workshop on Recommendation in Complex Scenarios (ComplexRec 2017) (Vol. 1892, pp. 23-28). CEUR-WS.

Matrix Factorization for Package Recommendations. / Wibowo, Agung Toto; Siddharthan, Advaith; Lin, Chenghua; Masthoff, Judith.

Proceedings of the RecSys 2017 Workshop on Recommendation in Complex Scenarios (ComplexRec 2017). ed. / Toine Bogers; Marijn Koolen; Bamshad Mobasher; Alan Said; Alexander Tuzhilin. Vol. 1892 CEUR-WS, 2017. p. 23-28.

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

Wibowo, AT, Siddharthan, A, Lin, C & Masthoff, J 2017, Matrix Factorization for Package Recommendations. in T Bogers, M Koolen, B Mobasher, A Said & A Tuzhilin (eds), Proceedings of the RecSys 2017 Workshop on Recommendation in Complex Scenarios (ComplexRec 2017). vol. 1892, CEUR-WS, pp. 23-28, ComplexRec 2017 Workshop, Como, Italy, 31/08/17.
Wibowo AT, Siddharthan A, Lin C, Masthoff J. Matrix Factorization for Package Recommendations. In Bogers T, Koolen M, Mobasher B, Said A, Tuzhilin A, editors, Proceedings of the RecSys 2017 Workshop on Recommendation in Complex Scenarios (ComplexRec 2017). Vol. 1892. CEUR-WS. 2017. p. 23-28
Wibowo, Agung Toto ; Siddharthan, Advaith ; Lin, Chenghua ; Masthoff, Judith. / Matrix Factorization for Package Recommendations. Proceedings of the RecSys 2017 Workshop on Recommendation in Complex Scenarios (ComplexRec 2017). editor / Toine Bogers ; Marijn Koolen ; Bamshad Mobasher ; Alan Said ; Alexander Tuzhilin. Vol. 1892 CEUR-WS, 2017. pp. 23-28
@inproceedings{86ab96b5e7bd41709e444b0e084bdaf4,
title = "Matrix Factorization for Package Recommendations",
abstract = "Research in recommendation systems has to date focused on recommending individual items to users. However there are contexts in which combinations of items need to be recommended, and there has been less research to date on how collaborative methods such as matrix factorization can be applied to such tasks. The research contributions of this paper are threefold. First, we formalize the collaborative package recommendation task as an extension of the standard collaborative recommendation task. Second, we describe and make available a novel package recommendation dataset in the clothes domain, where a combination of a {"}top{"} (e.g. a shirt, t-shirt or top) and {"}bottom'{"} (e.g. trousers, shorts or skirts) needs to be recommended. Finally, we describe several extensions of matrix factorization to predict user ratings on packages, and report RMSE improvements over the standard matrix factorization approach for recommending combinations of tops and bottoms.",
keywords = "Package Recommendation, Matrix Factorization, Clothes Domain, Collaborative Filtering",
author = "Wibowo, {Agung Toto} and Advaith Siddharthan and Chenghua Lin and Judith Masthoff",
note = "ACKNOWLEDGMENTS We would like to thank Lembaga Pengelola Dana Pendidikan (LPDP), Departemen Keuangan Indonesia for awarding a scholarship to support the studies of the lead author.",
year = "2017",
month = "8",
day = "18",
language = "English",
volume = "1892",
pages = "23--28",
editor = "Toine Bogers and Marijn Koolen and Bamshad Mobasher and Alan Said and Alexander Tuzhilin",
booktitle = "Proceedings of the RecSys 2017 Workshop on Recommendation in Complex Scenarios (ComplexRec 2017)",
publisher = "CEUR-WS",

}

TY - GEN

T1 - Matrix Factorization for Package Recommendations

AU - Wibowo, Agung Toto

AU - Siddharthan, Advaith

AU - Lin, Chenghua

AU - Masthoff, Judith

N1 - ACKNOWLEDGMENTS We would like to thank Lembaga Pengelola Dana Pendidikan (LPDP), Departemen Keuangan Indonesia for awarding a scholarship to support the studies of the lead author.

PY - 2017/8/18

Y1 - 2017/8/18

N2 - Research in recommendation systems has to date focused on recommending individual items to users. However there are contexts in which combinations of items need to be recommended, and there has been less research to date on how collaborative methods such as matrix factorization can be applied to such tasks. The research contributions of this paper are threefold. First, we formalize the collaborative package recommendation task as an extension of the standard collaborative recommendation task. Second, we describe and make available a novel package recommendation dataset in the clothes domain, where a combination of a "top" (e.g. a shirt, t-shirt or top) and "bottom'" (e.g. trousers, shorts or skirts) needs to be recommended. Finally, we describe several extensions of matrix factorization to predict user ratings on packages, and report RMSE improvements over the standard matrix factorization approach for recommending combinations of tops and bottoms.

AB - Research in recommendation systems has to date focused on recommending individual items to users. However there are contexts in which combinations of items need to be recommended, and there has been less research to date on how collaborative methods such as matrix factorization can be applied to such tasks. The research contributions of this paper are threefold. First, we formalize the collaborative package recommendation task as an extension of the standard collaborative recommendation task. Second, we describe and make available a novel package recommendation dataset in the clothes domain, where a combination of a "top" (e.g. a shirt, t-shirt or top) and "bottom'" (e.g. trousers, shorts or skirts) needs to be recommended. Finally, we describe several extensions of matrix factorization to predict user ratings on packages, and report RMSE improvements over the standard matrix factorization approach for recommending combinations of tops and bottoms.

KW - Package Recommendation

KW - Matrix Factorization

KW - Clothes Domain

KW - Collaborative Filtering

M3 - Conference contribution

VL - 1892

SP - 23

EP - 28

BT - Proceedings of the RecSys 2017 Workshop on Recommendation in Complex Scenarios (ComplexRec 2017)

A2 - Bogers, Toine

A2 - Koolen, Marijn

A2 - Mobasher, Bamshad

A2 - Said, Alan

A2 - Tuzhilin, Alexander

PB - CEUR-WS

ER -