Convolutional Adversarial Latent Factor Model for Recommender System

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

Abstract

The accuracy of Top-N recommendation task is challenged in the systems with mainly implicit user feedback considered. Adversarial training has presented successful results in identifying real data distributions in various domains (e.g. image processing). Nonetheless, adversarial training applied to recommendation is still challenged especially by interpretation of negative implicit feedback causing it to converge slowly as well as affecting its convergence stability. This is often attributed to high sparsity of the implicit feedback and discrete values characteristic from items recommendation. To face these challenges, we propose a novel model named convolutional adversarial latent factor model (CALF), which uses adversarial training in generative and discriminative models for implicit feedback recommendations. We assume that users prefer observed items over generated items and then apply pairwise product to model the user-item interactions. Additionally, the latent features become input data of our convolutional neural network (CNN) to learn correlations among embedding dimensions. Finally, Rao-Blackwellized sampling is adopted to deal with the discrete values optimizing CALF and stabilizing the training step. We conducted extensive experiments on three different benchmark datasets, where our proposed model demonstrates its efficiency for item recommendation.
Original languageEnglish
Title of host publicationProceedings of the Thirty-Second International Florida Artificial Intelligence Research Society Conference, Sarasota, Florida, USA, May 19-22 2019.
EditorsRoman Barták, Keith W. Brawner
Number of pages6
PublisherAAAI Press
Publication date2019
Pages419-424
Publication statusPublished - 2019
EventThe Thirty-Second International Flairs Conference - Lido Beach Resort, Florida, United States
Duration: 19 May 201922 May 2019
Conference number: Thirty-Second
https://aaai.org/ocs/index.php/FLAIRS/FLAIRS19/index

Conference

ConferenceThe Thirty-Second International Flairs Conference
NumberThirty-Second
LocationLido Beach Resort
CountryUnited States
CityFlorida
Period19/05/201922/05/2019
Internet address

Fingerprint

Recommender systems
Feedback
Image processing
Sampling
Neural networks
Experiments

Cite this

Costa, F. S. D., & Dolog, P. (2019). Convolutional Adversarial Latent Factor Model for Recommender System. In R. Barták, & K. W. Brawner (Eds.), Proceedings of the Thirty-Second International Florida Artificial Intelligence Research Society Conference, Sarasota, Florida, USA, May 19-22 2019. (pp. 419-424). AAAI Press.
Costa, Felipe Soares Da ; Dolog, Peter. / Convolutional Adversarial Latent Factor Model for Recommender System. Proceedings of the Thirty-Second International Florida Artificial Intelligence Research Society Conference, Sarasota, Florida, USA, May 19-22 2019.. editor / Roman Barták ; Keith W. Brawner. AAAI Press, 2019. pp. 419-424
@inproceedings{430678aa0a534e80b25ca1ffb35bb4cf,
title = "Convolutional Adversarial Latent Factor Model for Recommender System",
abstract = "The accuracy of Top-N recommendation task is challenged in the systems with mainly implicit user feedback considered. Adversarial training has presented successful results in identifying real data distributions in various domains (e.g. image processing). Nonetheless, adversarial training applied to recommendation is still challenged especially by interpretation of negative implicit feedback causing it to converge slowly as well as affecting its convergence stability. This is often attributed to high sparsity of the implicit feedback and discrete values characteristic from items recommendation. To face these challenges, we propose a novel model named convolutional adversarial latent factor model (CALF), which uses adversarial training in generative and discriminative models for implicit feedback recommendations. We assume that users prefer observed items over generated items and then apply pairwise product to model the user-item interactions. Additionally, the latent features become input data of our convolutional neural network (CNN) to learn correlations among embedding dimensions. Finally, Rao-Blackwellized sampling is adopted to deal with the discrete values optimizing CALF and stabilizing the training step. We conducted extensive experiments on three different benchmark datasets, where our proposed model demonstrates its efficiency for item recommendation.",
author = "Costa, {Felipe Soares Da} and Peter Dolog",
year = "2019",
language = "English",
pages = "419--424",
editor = "Roman Bart{\'a}k and Brawner, {Keith W.}",
booktitle = "Proceedings of the Thirty-Second International Florida Artificial Intelligence Research Society Conference, Sarasota, Florida, USA, May 19-22 2019.",
publisher = "AAAI Press",
address = "United States",

}

Costa, FSD & Dolog, P 2019, Convolutional Adversarial Latent Factor Model for Recommender System. in R Barták & KW Brawner (eds), Proceedings of the Thirty-Second International Florida Artificial Intelligence Research Society Conference, Sarasota, Florida, USA, May 19-22 2019.. AAAI Press, pp. 419-424, The Thirty-Second International Flairs Conference, Florida, United States, 19/05/2019.

Convolutional Adversarial Latent Factor Model for Recommender System. / Costa, Felipe Soares Da; Dolog, Peter.

Proceedings of the Thirty-Second International Florida Artificial Intelligence Research Society Conference, Sarasota, Florida, USA, May 19-22 2019.. ed. / Roman Barták; Keith W. Brawner. AAAI Press, 2019. p. 419-424.

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

TY - GEN

T1 - Convolutional Adversarial Latent Factor Model for Recommender System

AU - Costa, Felipe Soares Da

AU - Dolog, Peter

PY - 2019

Y1 - 2019

N2 - The accuracy of Top-N recommendation task is challenged in the systems with mainly implicit user feedback considered. Adversarial training has presented successful results in identifying real data distributions in various domains (e.g. image processing). Nonetheless, adversarial training applied to recommendation is still challenged especially by interpretation of negative implicit feedback causing it to converge slowly as well as affecting its convergence stability. This is often attributed to high sparsity of the implicit feedback and discrete values characteristic from items recommendation. To face these challenges, we propose a novel model named convolutional adversarial latent factor model (CALF), which uses adversarial training in generative and discriminative models for implicit feedback recommendations. We assume that users prefer observed items over generated items and then apply pairwise product to model the user-item interactions. Additionally, the latent features become input data of our convolutional neural network (CNN) to learn correlations among embedding dimensions. Finally, Rao-Blackwellized sampling is adopted to deal with the discrete values optimizing CALF and stabilizing the training step. We conducted extensive experiments on three different benchmark datasets, where our proposed model demonstrates its efficiency for item recommendation.

AB - The accuracy of Top-N recommendation task is challenged in the systems with mainly implicit user feedback considered. Adversarial training has presented successful results in identifying real data distributions in various domains (e.g. image processing). Nonetheless, adversarial training applied to recommendation is still challenged especially by interpretation of negative implicit feedback causing it to converge slowly as well as affecting its convergence stability. This is often attributed to high sparsity of the implicit feedback and discrete values characteristic from items recommendation. To face these challenges, we propose a novel model named convolutional adversarial latent factor model (CALF), which uses adversarial training in generative and discriminative models for implicit feedback recommendations. We assume that users prefer observed items over generated items and then apply pairwise product to model the user-item interactions. Additionally, the latent features become input data of our convolutional neural network (CNN) to learn correlations among embedding dimensions. Finally, Rao-Blackwellized sampling is adopted to deal with the discrete values optimizing CALF and stabilizing the training step. We conducted extensive experiments on three different benchmark datasets, where our proposed model demonstrates its efficiency for item recommendation.

M3 - Article in proceeding

SP - 419

EP - 424

BT - Proceedings of the Thirty-Second International Florida Artificial Intelligence Research Society Conference, Sarasota, Florida, USA, May 19-22 2019.

A2 - Barták, Roman

A2 - Brawner, Keith W.

PB - AAAI Press

ER -

Costa FSD, Dolog P. Convolutional Adversarial Latent Factor Model for Recommender System. In Barták R, Brawner KW, editors, Proceedings of the Thirty-Second International Florida Artificial Intelligence Research Society Conference, Sarasota, Florida, USA, May 19-22 2019.. AAAI Press. 2019. p. 419-424