Learning Manifold Representation from Multimodal Data for Event Detection in Flickr-Like Social Media

Zhenguo Yang; Qing Li; Wenyin Liu; Yun MAa

doi:10.1007/978-3-319-32055-7_14

Learning Manifold Representation from Multimodal Data for Event Detection in Flickr-Like Social Media

Zhenguo Yang^*
, Qing Li
, Wenyin Liu
, Yun MAa

^*Corresponding author for this work

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

3 Citations (Scopus)

Abstract

In this work, a three-stage social event detection model is devised to discover events in Flickr data. As the features possessed by the data are typically heterogeneous, a multimodal fusion model (M²F) exploits a soft-voting strategy and a reinforcing model is devised to learn fused features in the first stage. Furthermore, a Laplacian non-negative matrix factorization (LNMF) model is exploited to extract compact manifold representation. Particularly, a Laplacian regularization term constructed on the multimodal features is introduced to keep the geometry structure of the data. Finally, clustering algorithms can be applied seamlessly in order to detect event clusters. Extensive experiments conducted on the real-world dataset reveal the M²F-LNMF-based approaches outperform the baselines.

Original language	English
Title of host publication	Database Systems for Advanced Applications
Subtitle of host publication	DASFAA 2016 International Workshops: BDMS, BDQM, MoI, and SeCoP, Dallas, TX, USA, April 16-19, 2016, Proceedings
Editors	Hong Gao, Jinho Kim, Yasushi Sakurai
Publisher	Springer
Pages	160-167
ISBN (Electronic)	978-3-319-32055-7
ISBN (Print)	978-3-319-32054-0
DOIs	https://doi.org/10.1007/978-3-319-32055-7_14
Publication status	Published - 2016
Event	The 21st International Conference on Database Systems for Advanced Applications, DASFAA 2016 - The University of Texas at Dallas, Dallas, United States Duration: 16 Apr 2016 → 19 Apr 2016 http://theory.utdallas.edu/DASFAA2016/

Publication series

Name	Lecture Notes in Computer Science
Publisher	Springer
Volume	LNCS 9645
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	The 21st International Conference on Database Systems for Advanced Applications, DASFAA 2016
Abbreviated title	DASFAA 2016
Place	United States
City	Dallas
Period	16/04/16 → 19/04/16
Internet address	http://theory.utdallas.edu/DASFAA2016/

Research Keywords

Social media analytics
Multimedia content analysis
Multimodal fusion
Manifold learning
Event detection

Access Output

10.1007/978-3-319-32055-7_14

Other Access Options

Check CityUHK Library

Permanent Link

Right click to copy link

Cite this

Yang, Z., Li, Q., Liu, W., & MAa, Y. (2016). Learning Manifold Representation from Multimodal Data for Event Detection in Flickr-Like Social Media. In H. Gao, J. Kim, & Y. Sakurai (Eds.), Database Systems for Advanced Applications: DASFAA 2016 International Workshops: BDMS, BDQM, MoI, and SeCoP, Dallas, TX, USA, April 16-19, 2016, Proceedings (pp. 160-167). (Lecture Notes in Computer Science; Vol. LNCS 9645). Springer . https://doi.org/10.1007/978-3-319-32055-7_14

Yang, Zhenguo ; Li, Qing ; Liu, Wenyin et al. / Learning Manifold Representation from Multimodal Data for Event Detection in Flickr-Like Social Media. Database Systems for Advanced Applications: DASFAA 2016 International Workshops: BDMS, BDQM, MoI, and SeCoP, Dallas, TX, USA, April 16-19, 2016, Proceedings. editor / Hong Gao ; Jinho Kim ; Yasushi Sakurai. Springer , 2016. pp. 160-167 (Lecture Notes in Computer Science).

@inproceedings{c722f1fe5461414c98b91e918b93d911,

title = "Learning Manifold Representation from Multimodal Data for Event Detection in Flickr-Like Social Media",

abstract = "In this work, a three-stage social event detection model is devised to discover events in Flickr data. As the features possessed by the data are typically heterogeneous, a multimodal fusion model (M2F) exploits a soft-voting strategy and a reinforcing model is devised to learn fused features in the first stage. Furthermore, a Laplacian non-negative matrix factorization (LNMF) model is exploited to extract compact manifold representation. Particularly, a Laplacian regularization term constructed on the multimodal features is introduced to keep the geometry structure of the data. Finally, clustering algorithms can be applied seamlessly in order to detect event clusters. Extensive experiments conducted on the real-world dataset reveal the M2 F-LNMF-based approaches outperform the baselines.",

keywords = "Social media analytics, Multimedia content analysis, Multimodal fusion, Manifold learning, Event detection",

author = "Zhenguo Yang and Qing Li and Wenyin Liu and Yun MAa",

year = "2016",

doi = "10.1007/978-3-319-32055-7\_14",

language = "English",

isbn = "978-3-319-32054-0",

series = "Lecture Notes in Computer Science",

publisher = "Springer ",

pages = "160--167",

editor = "Hong Gao and Jinho Kim and Yasushi Sakurai",

booktitle = "Database Systems for Advanced Applications",

note = "The 21st International Conference on Database Systems for Advanced Applications, DASFAA 2016, DASFAA 2016 ; Conference date: 16-04-2016 Through 19-04-2016",

url = "http://theory.utdallas.edu/DASFAA2016/",

}

Yang, Z, Li, Q, Liu, W & MAa, Y 2016, Learning Manifold Representation from Multimodal Data for Event Detection in Flickr-Like Social Media. in H Gao, J Kim & Y Sakurai (eds), Database Systems for Advanced Applications: DASFAA 2016 International Workshops: BDMS, BDQM, MoI, and SeCoP, Dallas, TX, USA, April 16-19, 2016, Proceedings. Lecture Notes in Computer Science, vol. LNCS 9645, Springer , pp. 160-167, The 21st International Conference on Database Systems for Advanced Applications, DASFAA 2016, Dallas, Tennessee, United States, 16/04/16. https://doi.org/10.1007/978-3-319-32055-7_14

Learning Manifold Representation from Multimodal Data for Event Detection in Flickr-Like Social Media. / Yang, Zhenguo; Li, Qing; Liu, Wenyin et al.
Database Systems for Advanced Applications: DASFAA 2016 International Workshops: BDMS, BDQM, MoI, and SeCoP, Dallas, TX, USA, April 16-19, 2016, Proceedings. ed. / Hong Gao; Jinho Kim; Yasushi Sakurai. Springer , 2016. p. 160-167 (Lecture Notes in Computer Science; Vol. LNCS 9645).

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

TY - GEN

T1 - Learning Manifold Representation from Multimodal Data for Event Detection in Flickr-Like Social Media

AU - Yang, Zhenguo

AU - Li, Qing

AU - Liu, Wenyin

AU - MAa, Yun

PY - 2016

Y1 - 2016

N2 - In this work, a three-stage social event detection model is devised to discover events in Flickr data. As the features possessed by the data are typically heterogeneous, a multimodal fusion model (M2F) exploits a soft-voting strategy and a reinforcing model is devised to learn fused features in the first stage. Furthermore, a Laplacian non-negative matrix factorization (LNMF) model is exploited to extract compact manifold representation. Particularly, a Laplacian regularization term constructed on the multimodal features is introduced to keep the geometry structure of the data. Finally, clustering algorithms can be applied seamlessly in order to detect event clusters. Extensive experiments conducted on the real-world dataset reveal the M2 F-LNMF-based approaches outperform the baselines.

AB - In this work, a three-stage social event detection model is devised to discover events in Flickr data. As the features possessed by the data are typically heterogeneous, a multimodal fusion model (M2F) exploits a soft-voting strategy and a reinforcing model is devised to learn fused features in the first stage. Furthermore, a Laplacian non-negative matrix factorization (LNMF) model is exploited to extract compact manifold representation. Particularly, a Laplacian regularization term constructed on the multimodal features is introduced to keep the geometry structure of the data. Finally, clustering algorithms can be applied seamlessly in order to detect event clusters. Extensive experiments conducted on the real-world dataset reveal the M2 F-LNMF-based approaches outperform the baselines.

KW - Social media analytics

KW - Multimedia content analysis

KW - Multimodal fusion

KW - Manifold learning

KW - Event detection

UR - https://www.scopus.com/pages/publications/84964088560

UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-84964088560&origin=recordpage

U2 - 10.1007/978-3-319-32055-7_14

DO - 10.1007/978-3-319-32055-7_14

M3 - RGC 32 - Refereed conference paper (with host publication)

SN - 978-3-319-32054-0

T3 - Lecture Notes in Computer Science

SP - 160

EP - 167

BT - Database Systems for Advanced Applications

A2 - Gao, Hong

A2 - Kim, Jinho

A2 - Sakurai, Yasushi

PB - Springer

T2 - The 21st International Conference on Database Systems for Advanced Applications, DASFAA 2016

Y2 - 16 April 2016 through 19 April 2016

ER -

Yang Z, Li Q, Liu W, MAa Y. Learning Manifold Representation from Multimodal Data for Event Detection in Flickr-Like Social Media. In Gao H, Kim J, Sakurai Y, editors, Database Systems for Advanced Applications: DASFAA 2016 International Workshops: BDMS, BDQM, MoI, and SeCoP, Dallas, TX, USA, April 16-19, 2016, Proceedings. Springer . 2016. p. 160-167. (Lecture Notes in Computer Science). Epub 2016 Apr 12. doi: 10.1007/978-3-319-32055-7_14

Learning Manifold Representation from Multimodal Data for Event Detection in Flickr-Like Social Media

Abstract

Publication series

Conference

Research Keywords

Access Output

Other Access Options

Permanent Link

Other Files and Links

Fingerprint

Cite this