Multimodal sparse linear integration for content-based item recommendation

Qiusha Zhu, Zhao Li, Haohong Wang, Yimin Yang, Mei-Ling Shyu

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

7 Citations (Scopus)

Abstract

Most content-based recommender systems focus on analyzing the textual information of items. For items with images, the images can be treated as another information modality. In this paper, an effective method called MSLIM is proposed to integrate multimodal information for content-based item recommendation. It formalizes the probelm into a regularized optimization problem in the least-squares sense and the coordinate gradient descent is applied to solve the problem. The aggregation coefficients of the items are learned in an unsupervised manner during this process, based on which the k-nearest neighbor (k-NN) algorithm is used to generate the top-N recommendations of each item by finding its k nearest neighbors. A framework of using MSLIM for item recommendation is proposed accordingly. The experimental results on a self-collected handbag dataset show that MSLIM outperforms the selected comparison methods and show how the model parameters affect the final recommendation results.

Original languageEnglish
Title of host publicationProceedings - 2013 IEEE International Symposium on Multimedia, ISM 2013
Pages187-194
Number of pages8
DOIs
StatePublished - Dec 1 2013
Event15th IEEE International Symposium on Multimedia, ISM 2013 - Anaheim, CA, United States
Duration: Dec 9 2013Dec 11 2013

Other

Other15th IEEE International Symposium on Multimedia, ISM 2013
CountryUnited States
CityAnaheim, CA
Period12/9/1312/11/13

Fingerprint

Recommender systems
Agglomeration

Keywords

  • multimodal integration
  • Recommendation
  • sparse linear

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Human-Computer Interaction
  • Software

Cite this

Zhu, Q., Li, Z., Wang, H., Yang, Y., & Shyu, M-L. (2013). Multimodal sparse linear integration for content-based item recommendation. In Proceedings - 2013 IEEE International Symposium on Multimedia, ISM 2013 (pp. 187-194). [6746789] https://doi.org/10.1109/ISM.2013.37

Multimodal sparse linear integration for content-based item recommendation. / Zhu, Qiusha; Li, Zhao; Wang, Haohong; Yang, Yimin; Shyu, Mei-Ling.

Proceedings - 2013 IEEE International Symposium on Multimedia, ISM 2013. 2013. p. 187-194 6746789.

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

Zhu, Q, Li, Z, Wang, H, Yang, Y & Shyu, M-L 2013, Multimodal sparse linear integration for content-based item recommendation. in Proceedings - 2013 IEEE International Symposium on Multimedia, ISM 2013., 6746789, pp. 187-194, 15th IEEE International Symposium on Multimedia, ISM 2013, Anaheim, CA, United States, 12/9/13. https://doi.org/10.1109/ISM.2013.37
Zhu Q, Li Z, Wang H, Yang Y, Shyu M-L. Multimodal sparse linear integration for content-based item recommendation. In Proceedings - 2013 IEEE International Symposium on Multimedia, ISM 2013. 2013. p. 187-194. 6746789 https://doi.org/10.1109/ISM.2013.37
Zhu, Qiusha ; Li, Zhao ; Wang, Haohong ; Yang, Yimin ; Shyu, Mei-Ling. / Multimodal sparse linear integration for content-based item recommendation. Proceedings - 2013 IEEE International Symposium on Multimedia, ISM 2013. 2013. pp. 187-194
@inproceedings{cdf67fb7bf0d4c83885f2260b87885a3,
title = "Multimodal sparse linear integration for content-based item recommendation",
abstract = "Most content-based recommender systems focus on analyzing the textual information of items. For items with images, the images can be treated as another information modality. In this paper, an effective method called MSLIM is proposed to integrate multimodal information for content-based item recommendation. It formalizes the probelm into a regularized optimization problem in the least-squares sense and the coordinate gradient descent is applied to solve the problem. The aggregation coefficients of the items are learned in an unsupervised manner during this process, based on which the k-nearest neighbor (k-NN) algorithm is used to generate the top-N recommendations of each item by finding its k nearest neighbors. A framework of using MSLIM for item recommendation is proposed accordingly. The experimental results on a self-collected handbag dataset show that MSLIM outperforms the selected comparison methods and show how the model parameters affect the final recommendation results.",
keywords = "multimodal integration, Recommendation, sparse linear",
author = "Qiusha Zhu and Zhao Li and Haohong Wang and Yimin Yang and Mei-Ling Shyu",
year = "2013",
month = "12",
day = "1",
doi = "10.1109/ISM.2013.37",
language = "English",
isbn = "9780769551401",
pages = "187--194",
booktitle = "Proceedings - 2013 IEEE International Symposium on Multimedia, ISM 2013",

}

TY - GEN

T1 - Multimodal sparse linear integration for content-based item recommendation

AU - Zhu, Qiusha

AU - Li, Zhao

AU - Wang, Haohong

AU - Yang, Yimin

AU - Shyu, Mei-Ling

PY - 2013/12/1

Y1 - 2013/12/1

N2 - Most content-based recommender systems focus on analyzing the textual information of items. For items with images, the images can be treated as another information modality. In this paper, an effective method called MSLIM is proposed to integrate multimodal information for content-based item recommendation. It formalizes the probelm into a regularized optimization problem in the least-squares sense and the coordinate gradient descent is applied to solve the problem. The aggregation coefficients of the items are learned in an unsupervised manner during this process, based on which the k-nearest neighbor (k-NN) algorithm is used to generate the top-N recommendations of each item by finding its k nearest neighbors. A framework of using MSLIM for item recommendation is proposed accordingly. The experimental results on a self-collected handbag dataset show that MSLIM outperforms the selected comparison methods and show how the model parameters affect the final recommendation results.

AB - Most content-based recommender systems focus on analyzing the textual information of items. For items with images, the images can be treated as another information modality. In this paper, an effective method called MSLIM is proposed to integrate multimodal information for content-based item recommendation. It formalizes the probelm into a regularized optimization problem in the least-squares sense and the coordinate gradient descent is applied to solve the problem. The aggregation coefficients of the items are learned in an unsupervised manner during this process, based on which the k-nearest neighbor (k-NN) algorithm is used to generate the top-N recommendations of each item by finding its k nearest neighbors. A framework of using MSLIM for item recommendation is proposed accordingly. The experimental results on a self-collected handbag dataset show that MSLIM outperforms the selected comparison methods and show how the model parameters affect the final recommendation results.

KW - multimodal integration

KW - Recommendation

KW - sparse linear

UR - http://www.scopus.com/inward/record.url?scp=84900652045&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84900652045&partnerID=8YFLogxK

U2 - 10.1109/ISM.2013.37

DO - 10.1109/ISM.2013.37

M3 - Conference contribution

SN - 9780769551401

SP - 187

EP - 194

BT - Proceedings - 2013 IEEE International Symposium on Multimedia, ISM 2013

ER -