Leveraging concept association network for multimedia rare concept mining and retrieval

Tao Meng, Mei-Ling Shyu

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

30 Citations (Scopus)

Abstract

Automatic high-level semantic concept detection is a crucial step for multimedia data management, indexing, and retrieval. It is well-acknowledged that semantic gap poses a great challenge in multimedia content-based research. It becomes even more challenging when the concept of interest is extremely rare in the training data sets because of the poor modeling for the positive instances. In this paper, a Concept Association Network (CAN) is trained by selecting significant links to capture the strong associations among different concepts using association rule mining (ARM). By taking into account of the correlations and credibilities of reference concept nodes, the advantages of the reference nodes are utilized. Experimental results using TRECVID 2010 data sets show that by utilizing the proposed framework, the Mean Average Precision (MAP) values of all the concepts are improved, and the significant improvement of the MAP values of the rare concepts further attests the promising results.

Original languageEnglish (US)
Title of host publicationProceedings - IEEE International Conference on Multimedia and Expo
Pages860-865
Number of pages6
DOIs
StatePublished - 2012
Event2012 13th IEEE International Conference on Multimedia and Expo, ICME 2012 - Melbourne, VIC, Australia
Duration: Jul 9 2012Jul 13 2012

Other

Other2012 13th IEEE International Conference on Multimedia and Expo, ICME 2012
CountryAustralia
CityMelbourne, VIC
Period7/9/127/13/12

Fingerprint

Semantics
Association rules
Information management

Keywords

  • concept association network
  • Content-based multimedia retrieval
  • logistic regression
  • rare concept detection

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Computer Science Applications

Cite this

Meng, T., & Shyu, M-L. (2012). Leveraging concept association network for multimedia rare concept mining and retrieval. In Proceedings - IEEE International Conference on Multimedia and Expo (pp. 860-865). [6298511] https://doi.org/10.1109/ICME.2012.134

Leveraging concept association network for multimedia rare concept mining and retrieval. / Meng, Tao; Shyu, Mei-Ling.

Proceedings - IEEE International Conference on Multimedia and Expo. 2012. p. 860-865 6298511.

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

Meng, T & Shyu, M-L 2012, Leveraging concept association network for multimedia rare concept mining and retrieval. in Proceedings - IEEE International Conference on Multimedia and Expo., 6298511, pp. 860-865, 2012 13th IEEE International Conference on Multimedia and Expo, ICME 2012, Melbourne, VIC, Australia, 7/9/12. https://doi.org/10.1109/ICME.2012.134
Meng T, Shyu M-L. Leveraging concept association network for multimedia rare concept mining and retrieval. In Proceedings - IEEE International Conference on Multimedia and Expo. 2012. p. 860-865. 6298511 https://doi.org/10.1109/ICME.2012.134
Meng, Tao ; Shyu, Mei-Ling. / Leveraging concept association network for multimedia rare concept mining and retrieval. Proceedings - IEEE International Conference on Multimedia and Expo. 2012. pp. 860-865
@inproceedings{5c0579016aa6435b8efceb1f405e2dad,
title = "Leveraging concept association network for multimedia rare concept mining and retrieval",
abstract = "Automatic high-level semantic concept detection is a crucial step for multimedia data management, indexing, and retrieval. It is well-acknowledged that semantic gap poses a great challenge in multimedia content-based research. It becomes even more challenging when the concept of interest is extremely rare in the training data sets because of the poor modeling for the positive instances. In this paper, a Concept Association Network (CAN) is trained by selecting significant links to capture the strong associations among different concepts using association rule mining (ARM). By taking into account of the correlations and credibilities of reference concept nodes, the advantages of the reference nodes are utilized. Experimental results using TRECVID 2010 data sets show that by utilizing the proposed framework, the Mean Average Precision (MAP) values of all the concepts are improved, and the significant improvement of the MAP values of the rare concepts further attests the promising results.",
keywords = "concept association network, Content-based multimedia retrieval, logistic regression, rare concept detection",
author = "Tao Meng and Mei-Ling Shyu",
year = "2012",
doi = "10.1109/ICME.2012.134",
language = "English (US)",
pages = "860--865",
booktitle = "Proceedings - IEEE International Conference on Multimedia and Expo",

}

TY - GEN

T1 - Leveraging concept association network for multimedia rare concept mining and retrieval

AU - Meng, Tao

AU - Shyu, Mei-Ling

PY - 2012

Y1 - 2012

N2 - Automatic high-level semantic concept detection is a crucial step for multimedia data management, indexing, and retrieval. It is well-acknowledged that semantic gap poses a great challenge in multimedia content-based research. It becomes even more challenging when the concept of interest is extremely rare in the training data sets because of the poor modeling for the positive instances. In this paper, a Concept Association Network (CAN) is trained by selecting significant links to capture the strong associations among different concepts using association rule mining (ARM). By taking into account of the correlations and credibilities of reference concept nodes, the advantages of the reference nodes are utilized. Experimental results using TRECVID 2010 data sets show that by utilizing the proposed framework, the Mean Average Precision (MAP) values of all the concepts are improved, and the significant improvement of the MAP values of the rare concepts further attests the promising results.

AB - Automatic high-level semantic concept detection is a crucial step for multimedia data management, indexing, and retrieval. It is well-acknowledged that semantic gap poses a great challenge in multimedia content-based research. It becomes even more challenging when the concept of interest is extremely rare in the training data sets because of the poor modeling for the positive instances. In this paper, a Concept Association Network (CAN) is trained by selecting significant links to capture the strong associations among different concepts using association rule mining (ARM). By taking into account of the correlations and credibilities of reference concept nodes, the advantages of the reference nodes are utilized. Experimental results using TRECVID 2010 data sets show that by utilizing the proposed framework, the Mean Average Precision (MAP) values of all the concepts are improved, and the significant improvement of the MAP values of the rare concepts further attests the promising results.

KW - concept association network

KW - Content-based multimedia retrieval

KW - logistic regression

KW - rare concept detection

UR - http://www.scopus.com/inward/record.url?scp=84868135897&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84868135897&partnerID=8YFLogxK

U2 - 10.1109/ICME.2012.134

DO - 10.1109/ICME.2012.134

M3 - Conference contribution

SP - 860

EP - 865

BT - Proceedings - IEEE International Conference on Multimedia and Expo

ER -