Incorporating Constraints into Matrix Factorization for Clothes Package Recommendation

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

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

1 Citation (Scopus)

Abstract

Recommender systems have been widely applied in the literature to suggest individual items to users. In this paper, we consider the harder problem of package recommendation, where items are recommended together as a package. We focus on the clothing domain, where a package recommendation involves a combination of a 'top' (e.g. a shirt) and a 'bottom' (e.g. a pair of trousers). The novelty in this work is that we combined matrix factorisation methods for collaborative filtering with hand-crafted and learnt fashion constraints on combining item features such as colour, formality and patterns. Finally, to better understand where the algorithms are underperforming, we conducted focus groups, which lead to deeper insights into how to use constraints to improve package recommendation in this domain
Original languageEnglish
Title of host publicationProceedings of the 26th ACM Conference on User Modelling, Adaptation and Personalization (UMAP)
Place of PublicationSingapore
PublisherACM
Pages111-119
Number of pages9
ISBN (Print)978-1-4503-5589-6
Publication statusPublished - 2018
EventUser Modeling, Adaptation and Personalization - ACM UMAP 2018: User Modelling, Adaptation and Personalization - Nanyang Technological University, Singapore, Singapore
Duration: 8 Jul 201711 Jul 2018
http://www.um.org/umap2018/ (User Modeling, Adaptation and Personalization - ACM UMAP 2018)

Conference

ConferenceUser Modeling, Adaptation and Personalization - ACM UMAP 2018
CountrySingapore
CitySingapore
Period8/07/1711/07/18
Internet address

Fingerprint

Collaborative filtering
Recommender systems
Factorization
Color

Cite this

Wibowo, A. T., Siddharthan, A., Masthoff, J. F. M., & Lin, C. (2018). Incorporating Constraints into Matrix Factorization for Clothes Package Recommendation. In Proceedings of the 26th ACM Conference on User Modelling, Adaptation and Personalization (UMAP) (pp. 111-119). Singapore: ACM.

Incorporating Constraints into Matrix Factorization for Clothes Package Recommendation. / Wibowo, Agung Toto; Siddharthan, Advaith; Masthoff, Judith Francoise Maria; Lin, Chenghua.

Proceedings of the 26th ACM Conference on User Modelling, Adaptation and Personalization (UMAP). Singapore : ACM, 2018. p. 111-119.

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

Wibowo, AT, Siddharthan, A, Masthoff, JFM & Lin, C 2018, Incorporating Constraints into Matrix Factorization for Clothes Package Recommendation. in Proceedings of the 26th ACM Conference on User Modelling, Adaptation and Personalization (UMAP). ACM, Singapore, pp. 111-119, User Modeling, Adaptation and Personalization - ACM UMAP 2018, Singapore, Singapore, 8/07/17.
Wibowo AT, Siddharthan A, Masthoff JFM, Lin C. Incorporating Constraints into Matrix Factorization for Clothes Package Recommendation. In Proceedings of the 26th ACM Conference on User Modelling, Adaptation and Personalization (UMAP). Singapore: ACM. 2018. p. 111-119
Wibowo, Agung Toto ; Siddharthan, Advaith ; Masthoff, Judith Francoise Maria ; Lin, Chenghua. / Incorporating Constraints into Matrix Factorization for Clothes Package Recommendation. Proceedings of the 26th ACM Conference on User Modelling, Adaptation and Personalization (UMAP). Singapore : ACM, 2018. pp. 111-119
@inproceedings{079a354deafe4448b7fa7498189767a2,
title = "Incorporating Constraints into Matrix Factorization for Clothes Package Recommendation",
abstract = "Recommender systems have been widely applied in the literature to suggest individual items to users. In this paper, we consider the harder problem of package recommendation, where items are recommended together as a package. We focus on the clothing domain, where a package recommendation involves a combination of a 'top' (e.g. a shirt) and a 'bottom' (e.g. a pair of trousers). The novelty in this work is that we combined matrix factorisation methods for collaborative filtering with hand-crafted and learnt fashion constraints on combining item features such as colour, formality and patterns. Finally, to better understand where the algorithms are underperforming, we conducted focus groups, which lead to deeper insights into how to use constraints to improve package recommendation in this domain",
author = "Wibowo, {Agung Toto} and Advaith Siddharthan and Masthoff, {Judith Francoise Maria} and Chenghua Lin",
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. We would also like to thank the participants in our focus groups who communicated precious feedback during our discussions",
year = "2018",
language = "English",
isbn = "978-1-4503-5589-6",
pages = "111--119",
booktitle = "Proceedings of the 26th ACM Conference on User Modelling, Adaptation and Personalization (UMAP)",
publisher = "ACM",

}

TY - GEN

T1 - Incorporating Constraints into Matrix Factorization for Clothes Package Recommendation

AU - Wibowo, Agung Toto

AU - Siddharthan, Advaith

AU - Masthoff, Judith Francoise Maria

AU - Lin, Chenghua

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. We would also like to thank the participants in our focus groups who communicated precious feedback during our discussions

PY - 2018

Y1 - 2018

N2 - Recommender systems have been widely applied in the literature to suggest individual items to users. In this paper, we consider the harder problem of package recommendation, where items are recommended together as a package. We focus on the clothing domain, where a package recommendation involves a combination of a 'top' (e.g. a shirt) and a 'bottom' (e.g. a pair of trousers). The novelty in this work is that we combined matrix factorisation methods for collaborative filtering with hand-crafted and learnt fashion constraints on combining item features such as colour, formality and patterns. Finally, to better understand where the algorithms are underperforming, we conducted focus groups, which lead to deeper insights into how to use constraints to improve package recommendation in this domain

AB - Recommender systems have been widely applied in the literature to suggest individual items to users. In this paper, we consider the harder problem of package recommendation, where items are recommended together as a package. We focus on the clothing domain, where a package recommendation involves a combination of a 'top' (e.g. a shirt) and a 'bottom' (e.g. a pair of trousers). The novelty in this work is that we combined matrix factorisation methods for collaborative filtering with hand-crafted and learnt fashion constraints on combining item features such as colour, formality and patterns. Finally, to better understand where the algorithms are underperforming, we conducted focus groups, which lead to deeper insights into how to use constraints to improve package recommendation in this domain

M3 - Conference contribution

SN - 978-1-4503-5589-6

SP - 111

EP - 119

BT - Proceedings of the 26th ACM Conference on User Modelling, Adaptation and Personalization (UMAP)

PB - ACM

CY - Singapore

ER -