Sain: Self-attentive integration network for recommendation

Seoungjun Yun, Raehyun Kim, Miyoung Ko, Jaewoo Kang

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

1 Citation (Scopus)

Abstract

With the growing importance of personalized recommendation, numerous recommendation models have been proposed recently. Among them, Matrix Factorization (MF) based models are the most widely used in the recommendation field due to their high performance. However, MF based models suffer from cold start problems where user-item interactions are sparse. To deal with this problem, content based recommendation models which use the auxiliary attributes of users and items have been proposed. Since these models use auxiliary attributes, they are effective in cold start settings. However, most of the proposed models are either unable to capture complex feature interactions or not properly designed to combine user-item feedback information with content information. In this paper, we propose Self-Attentive Integration Network (SAIN) which is a model that effectively combines user-item feedback information and auxiliary information for recommendation task. In SAIN, a self-attention mechanism is used in the feature-level interaction layer to effectively consider interactions between multiple features, while the information integration layer adaptively combines content and feedback information. The experimental results on two public datasets show that our model outperforms the state-of-the-art models by 2.13%.

Original languageEnglish
Title of host publicationSIGIR 2019 - Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval
PublisherAssociation for Computing Machinery, Inc
Pages1205-1208
Number of pages4
ISBN (Electronic)9781450361729
DOIs
Publication statusPublished - 2019 Jul 18
Event42nd International ACM SIGIR Conference on Research and Development in Information Retrieval, SIGIR 2019 - Paris, France
Duration: 2019 Jul 212019 Jul 25

Publication series

NameSIGIR 2019 - Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval

Conference

Conference42nd International ACM SIGIR Conference on Research and Development in Information Retrieval, SIGIR 2019
CountryFrance
CityParis
Period19/7/2119/7/25

Fingerprint

Recommendations
Matrix Factorization
Model
Factorization
Feedback
Interaction
Attribute
Personalized Recommendation
Auxiliary Information
Information Integration
Information Content
High Performance
Experimental Results

Keywords

  • Datasets
  • Gaze detection
  • Neural networks
  • Text tagging

ASJC Scopus subject areas

  • Information Systems
  • Applied Mathematics
  • Software

Cite this

Yun, S., Kim, R., Ko, M., & Kang, J. (2019). Sain: Self-attentive integration network for recommendation. In SIGIR 2019 - Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval (pp. 1205-1208). (SIGIR 2019 - Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval). Association for Computing Machinery, Inc. https://doi.org/10.1145/3331184.3331342

Sain : Self-attentive integration network for recommendation. / Yun, Seoungjun; Kim, Raehyun; Ko, Miyoung; Kang, Jaewoo.

SIGIR 2019 - Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval. Association for Computing Machinery, Inc, 2019. p. 1205-1208 (SIGIR 2019 - Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval).

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

Yun, S, Kim, R, Ko, M & Kang, J 2019, Sain: Self-attentive integration network for recommendation. in SIGIR 2019 - Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval. SIGIR 2019 - Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval, Association for Computing Machinery, Inc, pp. 1205-1208, 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval, SIGIR 2019, Paris, France, 19/7/21. https://doi.org/10.1145/3331184.3331342
Yun S, Kim R, Ko M, Kang J. Sain: Self-attentive integration network for recommendation. In SIGIR 2019 - Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval. Association for Computing Machinery, Inc. 2019. p. 1205-1208. (SIGIR 2019 - Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval). https://doi.org/10.1145/3331184.3331342
Yun, Seoungjun ; Kim, Raehyun ; Ko, Miyoung ; Kang, Jaewoo. / Sain : Self-attentive integration network for recommendation. SIGIR 2019 - Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval. Association for Computing Machinery, Inc, 2019. pp. 1205-1208 (SIGIR 2019 - Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval).
@inproceedings{9bc12a7400ab407e82fa5146ed49fff7,
title = "Sain: Self-attentive integration network for recommendation",
abstract = "With the growing importance of personalized recommendation, numerous recommendation models have been proposed recently. Among them, Matrix Factorization (MF) based models are the most widely used in the recommendation field due to their high performance. However, MF based models suffer from cold start problems where user-item interactions are sparse. To deal with this problem, content based recommendation models which use the auxiliary attributes of users and items have been proposed. Since these models use auxiliary attributes, they are effective in cold start settings. However, most of the proposed models are either unable to capture complex feature interactions or not properly designed to combine user-item feedback information with content information. In this paper, we propose Self-Attentive Integration Network (SAIN) which is a model that effectively combines user-item feedback information and auxiliary information for recommendation task. In SAIN, a self-attention mechanism is used in the feature-level interaction layer to effectively consider interactions between multiple features, while the information integration layer adaptively combines content and feedback information. The experimental results on two public datasets show that our model outperforms the state-of-the-art models by 2.13{\%}.",
keywords = "Datasets, Gaze detection, Neural networks, Text tagging",
author = "Seoungjun Yun and Raehyun Kim and Miyoung Ko and Jaewoo Kang",
year = "2019",
month = "7",
day = "18",
doi = "10.1145/3331184.3331342",
language = "English",
series = "SIGIR 2019 - Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval",
publisher = "Association for Computing Machinery, Inc",
pages = "1205--1208",
booktitle = "SIGIR 2019 - Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval",

}

TY - GEN

T1 - Sain

T2 - Self-attentive integration network for recommendation

AU - Yun, Seoungjun

AU - Kim, Raehyun

AU - Ko, Miyoung

AU - Kang, Jaewoo

PY - 2019/7/18

Y1 - 2019/7/18

N2 - With the growing importance of personalized recommendation, numerous recommendation models have been proposed recently. Among them, Matrix Factorization (MF) based models are the most widely used in the recommendation field due to their high performance. However, MF based models suffer from cold start problems where user-item interactions are sparse. To deal with this problem, content based recommendation models which use the auxiliary attributes of users and items have been proposed. Since these models use auxiliary attributes, they are effective in cold start settings. However, most of the proposed models are either unable to capture complex feature interactions or not properly designed to combine user-item feedback information with content information. In this paper, we propose Self-Attentive Integration Network (SAIN) which is a model that effectively combines user-item feedback information and auxiliary information for recommendation task. In SAIN, a self-attention mechanism is used in the feature-level interaction layer to effectively consider interactions between multiple features, while the information integration layer adaptively combines content and feedback information. The experimental results on two public datasets show that our model outperforms the state-of-the-art models by 2.13%.

AB - With the growing importance of personalized recommendation, numerous recommendation models have been proposed recently. Among them, Matrix Factorization (MF) based models are the most widely used in the recommendation field due to their high performance. However, MF based models suffer from cold start problems where user-item interactions are sparse. To deal with this problem, content based recommendation models which use the auxiliary attributes of users and items have been proposed. Since these models use auxiliary attributes, they are effective in cold start settings. However, most of the proposed models are either unable to capture complex feature interactions or not properly designed to combine user-item feedback information with content information. In this paper, we propose Self-Attentive Integration Network (SAIN) which is a model that effectively combines user-item feedback information and auxiliary information for recommendation task. In SAIN, a self-attention mechanism is used in the feature-level interaction layer to effectively consider interactions between multiple features, while the information integration layer adaptively combines content and feedback information. The experimental results on two public datasets show that our model outperforms the state-of-the-art models by 2.13%.

KW - Datasets

KW - Gaze detection

KW - Neural networks

KW - Text tagging

UR - http://www.scopus.com/inward/record.url?scp=85073802746&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85073802746&partnerID=8YFLogxK

U2 - 10.1145/3331184.3331342

DO - 10.1145/3331184.3331342

M3 - Conference contribution

AN - SCOPUS:85073802746

T3 - SIGIR 2019 - Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval

SP - 1205

EP - 1208

BT - SIGIR 2019 - Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval

PB - Association for Computing Machinery, Inc

ER -