Context-aware LDA: Balancing Relevance and Diversity in TV Content Recommenders

Jing Yuan, Fikret Sivrikaya, Frank Hopfgartner, Andreas Lommatzsch, Mu Mu

Research output: Contribution to conference typesPaperResearchpeer-review

Abstract

In the vast and expanding ocean of digital content, users are hardly satisfied with recommended programs solely based on static user patterns and common statistics. Therefore, there is growing interest in recommendation approaches that aim to provide a certain level of diversity, besides precision and ranking. Context-awareness, which is an effective way to express dynamics and adaptivity, is widely used in recom-mender systems to set a proper balance between ranking and diversity. In light of these observations, we introduce a recommender with a context-aware probabilistic graphi-cal model and apply it to a campus-wide TV content de-livery system named “Vision”. Within this recommender, selection criteria of candidate fields and contextual factors are designed and users’ dependencies on their personal pref-erence or the aforementioned contextual influences can be distinguished. Most importantly, as to the role of balanc-ing relevance and diversity, final experiment results prove that context-aware LDA can evidently outperform other al-gorithms on both metrics. Thus this scalable model can be flexibly used for different recommendation purposes.
Original languageEnglish
Pages1-8
Number of pages8
Publication statusPublished - 19 Sep 2015
EventThe 9th ACM Conference on Recommender Systems - Vienna, Austria
Duration: 16 Sep 201520 Sep 2015

Conference

ConferenceThe 9th ACM Conference on Recommender Systems
CountryAustria
CityVienna
Period16/09/1520/09/15

Fingerprint

Statistics
Experiments

Bibliographical note

ISBN: 978-1-4503-3692-5

Keywords

  • TV recommender
  • context-aware
  • Latent Dirichlet Allocation
  • ranking
  • diversity

Cite this

Yuan, J., Sivrikaya, F., Hopfgartner, F., Lommatzsch, A., & Mu, M. (2015). Context-aware LDA: Balancing Relevance and Diversity in TV Content Recommenders. 1-8. Paper presented at The 9th ACM Conference on Recommender Systems , Vienna, Austria.
Yuan, Jing ; Sivrikaya, Fikret ; Hopfgartner, Frank ; Lommatzsch, Andreas ; Mu, Mu. / Context-aware LDA: Balancing Relevance and Diversity in TV Content Recommenders. Paper presented at The 9th ACM Conference on Recommender Systems , Vienna, Austria.8 p.
@conference{245356ccd3e645bd93349ac5f2f46cda,
title = "Context-aware LDA: Balancing Relevance and Diversity in TV Content Recommenders",
abstract = "In the vast and expanding ocean of digital content, users are hardly satisfied with recommended programs solely based on static user patterns and common statistics. Therefore, there is growing interest in recommendation approaches that aim to provide a certain level of diversity, besides precision and ranking. Context-awareness, which is an effective way to express dynamics and adaptivity, is widely used in recom-mender systems to set a proper balance between ranking and diversity. In light of these observations, we introduce a recommender with a context-aware probabilistic graphi-cal model and apply it to a campus-wide TV content de-livery system named “Vision”. Within this recommender, selection criteria of candidate fields and contextual factors are designed and users’ dependencies on their personal pref-erence or the aforementioned contextual influences can be distinguished. Most importantly, as to the role of balanc-ing relevance and diversity, final experiment results prove that context-aware LDA can evidently outperform other al-gorithms on both metrics. Thus this scalable model can be flexibly used for different recommendation purposes.",
keywords = "TV recommender, context-aware, Latent Dirichlet Allocation, ranking, diversity",
author = "Jing Yuan and Fikret Sivrikaya and Frank Hopfgartner and Andreas Lommatzsch and Mu Mu",
note = "ISBN: 978-1-4503-3692-5; The 9th ACM Conference on Recommender Systems ; Conference date: 16-09-2015 Through 20-09-2015",
year = "2015",
month = "9",
day = "19",
language = "English",
pages = "1--8",

}

Yuan, J, Sivrikaya, F, Hopfgartner, F, Lommatzsch, A & Mu, M 2015, 'Context-aware LDA: Balancing Relevance and Diversity in TV Content Recommenders' Paper presented at The 9th ACM Conference on Recommender Systems , Vienna, Austria, 16/09/15 - 20/09/15, pp. 1-8.

Context-aware LDA: Balancing Relevance and Diversity in TV Content Recommenders. / Yuan, Jing; Sivrikaya, Fikret; Hopfgartner, Frank; Lommatzsch, Andreas; Mu, Mu.

2015. 1-8 Paper presented at The 9th ACM Conference on Recommender Systems , Vienna, Austria.

Research output: Contribution to conference typesPaperResearchpeer-review

TY - CONF

T1 - Context-aware LDA: Balancing Relevance and Diversity in TV Content Recommenders

AU - Yuan, Jing

AU - Sivrikaya, Fikret

AU - Hopfgartner, Frank

AU - Lommatzsch, Andreas

AU - Mu, Mu

N1 - ISBN: 978-1-4503-3692-5

PY - 2015/9/19

Y1 - 2015/9/19

N2 - In the vast and expanding ocean of digital content, users are hardly satisfied with recommended programs solely based on static user patterns and common statistics. Therefore, there is growing interest in recommendation approaches that aim to provide a certain level of diversity, besides precision and ranking. Context-awareness, which is an effective way to express dynamics and adaptivity, is widely used in recom-mender systems to set a proper balance between ranking and diversity. In light of these observations, we introduce a recommender with a context-aware probabilistic graphi-cal model and apply it to a campus-wide TV content de-livery system named “Vision”. Within this recommender, selection criteria of candidate fields and contextual factors are designed and users’ dependencies on their personal pref-erence or the aforementioned contextual influences can be distinguished. Most importantly, as to the role of balanc-ing relevance and diversity, final experiment results prove that context-aware LDA can evidently outperform other al-gorithms on both metrics. Thus this scalable model can be flexibly used for different recommendation purposes.

AB - In the vast and expanding ocean of digital content, users are hardly satisfied with recommended programs solely based on static user patterns and common statistics. Therefore, there is growing interest in recommendation approaches that aim to provide a certain level of diversity, besides precision and ranking. Context-awareness, which is an effective way to express dynamics and adaptivity, is widely used in recom-mender systems to set a proper balance between ranking and diversity. In light of these observations, we introduce a recommender with a context-aware probabilistic graphi-cal model and apply it to a campus-wide TV content de-livery system named “Vision”. Within this recommender, selection criteria of candidate fields and contextual factors are designed and users’ dependencies on their personal pref-erence or the aforementioned contextual influences can be distinguished. Most importantly, as to the role of balanc-ing relevance and diversity, final experiment results prove that context-aware LDA can evidently outperform other al-gorithms on both metrics. Thus this scalable model can be flexibly used for different recommendation purposes.

KW - TV recommender

KW - context-aware

KW - Latent Dirichlet Allocation

KW - ranking

KW - diversity

M3 - Paper

SP - 1

EP - 8

ER -

Yuan J, Sivrikaya F, Hopfgartner F, Lommatzsch A, Mu M. Context-aware LDA: Balancing Relevance and Diversity in TV Content Recommenders. 2015. Paper presented at The 9th ACM Conference on Recommender Systems , Vienna, Austria.