Reliable TF-based recommender system for capturing complex correlations among contexts

Byungkook Oh, Sangjin Shin, Sungkwang Eom, Jooik Jung, Minjae Song, Seungmin Seo, Kyong Ho Lee

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Context-aware recommender systems (CARS) exploit multiple contexts to improve user experience in embracing new information and services. Tensor factorization (TF), a type of latent factor model, has achieved remarkable performance in CARS. TF learns latent representations of contexts by decomposing an observed rating tensor and combines the latent representations as a vector form to represent contextual influence on users and items. However, due to the limitation of the contextual expression power, they have difficulties in effectively capturing complex correlations among multiple contexts, and also the meaning of each context is diluted. To address the issue, we propose a reliable TF-based recommender system based on a proposed context tensor (CT-CARS), which incorporates a variety of correlations among contexts. CT-CARS contains a novel recommendation rating function and a learning algorithm. Specifically, the proposed context tensor elaborately captures the influences of both individual contexts and context combinations. Moreover, we introduce a novel parameter initialization based on past-learned results to improve the reliability of recommendations. The overall time complexity of our parameter learning algorithm grows linearly as dataset size increases. Experiments on six real-world datasets including two large-scaled datasets show that CT-CARS outperforms the existing state-of-the-art models in terms of both accuracy and reliability.

Original languageEnglish
Pages (from-to)337-365
Number of pages29
JournalJournal of Intelligent Information Systems
Volume52
Issue number2
DOIs
Publication statusPublished - 2019 Apr 15

Fingerprint

Recommender systems
Factorization
Tensors
Learning algorithms
Experiments

All Science Journal Classification (ASJC) codes

  • Software
  • Information Systems
  • Hardware and Architecture
  • Computer Networks and Communications
  • Artificial Intelligence

Cite this

Oh, Byungkook ; Shin, Sangjin ; Eom, Sungkwang ; Jung, Jooik ; Song, Minjae ; Seo, Seungmin ; Lee, Kyong Ho. / Reliable TF-based recommender system for capturing complex correlations among contexts. In: Journal of Intelligent Information Systems. 2019 ; Vol. 52, No. 2. pp. 337-365.
@article{91f0377c65d94960af3115ffbb459570,
title = "Reliable TF-based recommender system for capturing complex correlations among contexts",
abstract = "Context-aware recommender systems (CARS) exploit multiple contexts to improve user experience in embracing new information and services. Tensor factorization (TF), a type of latent factor model, has achieved remarkable performance in CARS. TF learns latent representations of contexts by decomposing an observed rating tensor and combines the latent representations as a vector form to represent contextual influence on users and items. However, due to the limitation of the contextual expression power, they have difficulties in effectively capturing complex correlations among multiple contexts, and also the meaning of each context is diluted. To address the issue, we propose a reliable TF-based recommender system based on a proposed context tensor (CT-CARS), which incorporates a variety of correlations among contexts. CT-CARS contains a novel recommendation rating function and a learning algorithm. Specifically, the proposed context tensor elaborately captures the influences of both individual contexts and context combinations. Moreover, we introduce a novel parameter initialization based on past-learned results to improve the reliability of recommendations. The overall time complexity of our parameter learning algorithm grows linearly as dataset size increases. Experiments on six real-world datasets including two large-scaled datasets show that CT-CARS outperforms the existing state-of-the-art models in terms of both accuracy and reliability.",
author = "Byungkook Oh and Sangjin Shin and Sungkwang Eom and Jooik Jung and Minjae Song and Seungmin Seo and Lee, {Kyong Ho}",
year = "2019",
month = "4",
day = "15",
doi = "10.1007/s10844-018-0514-7",
language = "English",
volume = "52",
pages = "337--365",
journal = "Journal of Intelligent Information Systems",
issn = "0925-9902",
publisher = "Springer Netherlands",
number = "2",

}

Reliable TF-based recommender system for capturing complex correlations among contexts. / Oh, Byungkook; Shin, Sangjin; Eom, Sungkwang; Jung, Jooik; Song, Minjae; Seo, Seungmin; Lee, Kyong Ho.

In: Journal of Intelligent Information Systems, Vol. 52, No. 2, 15.04.2019, p. 337-365.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Reliable TF-based recommender system for capturing complex correlations among contexts

AU - Oh, Byungkook

AU - Shin, Sangjin

AU - Eom, Sungkwang

AU - Jung, Jooik

AU - Song, Minjae

AU - Seo, Seungmin

AU - Lee, Kyong Ho

PY - 2019/4/15

Y1 - 2019/4/15

N2 - Context-aware recommender systems (CARS) exploit multiple contexts to improve user experience in embracing new information and services. Tensor factorization (TF), a type of latent factor model, has achieved remarkable performance in CARS. TF learns latent representations of contexts by decomposing an observed rating tensor and combines the latent representations as a vector form to represent contextual influence on users and items. However, due to the limitation of the contextual expression power, they have difficulties in effectively capturing complex correlations among multiple contexts, and also the meaning of each context is diluted. To address the issue, we propose a reliable TF-based recommender system based on a proposed context tensor (CT-CARS), which incorporates a variety of correlations among contexts. CT-CARS contains a novel recommendation rating function and a learning algorithm. Specifically, the proposed context tensor elaborately captures the influences of both individual contexts and context combinations. Moreover, we introduce a novel parameter initialization based on past-learned results to improve the reliability of recommendations. The overall time complexity of our parameter learning algorithm grows linearly as dataset size increases. Experiments on six real-world datasets including two large-scaled datasets show that CT-CARS outperforms the existing state-of-the-art models in terms of both accuracy and reliability.

AB - Context-aware recommender systems (CARS) exploit multiple contexts to improve user experience in embracing new information and services. Tensor factorization (TF), a type of latent factor model, has achieved remarkable performance in CARS. TF learns latent representations of contexts by decomposing an observed rating tensor and combines the latent representations as a vector form to represent contextual influence on users and items. However, due to the limitation of the contextual expression power, they have difficulties in effectively capturing complex correlations among multiple contexts, and also the meaning of each context is diluted. To address the issue, we propose a reliable TF-based recommender system based on a proposed context tensor (CT-CARS), which incorporates a variety of correlations among contexts. CT-CARS contains a novel recommendation rating function and a learning algorithm. Specifically, the proposed context tensor elaborately captures the influences of both individual contexts and context combinations. Moreover, we introduce a novel parameter initialization based on past-learned results to improve the reliability of recommendations. The overall time complexity of our parameter learning algorithm grows linearly as dataset size increases. Experiments on six real-world datasets including two large-scaled datasets show that CT-CARS outperforms the existing state-of-the-art models in terms of both accuracy and reliability.

UR - http://www.scopus.com/inward/record.url?scp=85049008041&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85049008041&partnerID=8YFLogxK

U2 - 10.1007/s10844-018-0514-7

DO - 10.1007/s10844-018-0514-7

M3 - Article

AN - SCOPUS:85049008041

VL - 52

SP - 337

EP - 365

JO - Journal of Intelligent Information Systems

JF - Journal of Intelligent Information Systems

SN - 0925-9902

IS - 2

ER -