Multi-graph enabled active learning for multimodal web image retrieval

Xin-Jing Wang; Xing Li; Wei-Ying Ma; Lei Zhang

doi:10.1145/1101826.1101839

Multi-graph enabled active learning for multimodal web image retrieval

Xin-Jing Wang, Xing Li, Wei-Ying Ma, Lei Zhang

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

5 Citations (Scopus)

Abstract

Tsinghua University, China In this paper, we propose a multimodal Web image retrieval technique based on multi-graph enabled active learning. The main goal is to leverage the heterogeneous data on the Web to improve retrieval precision. Three graphs are constructed on images' content features, textual annotations and hyperlinks respectively, namely Content-Graph, Text-Graph and Link-Graph, which provide complimentary information on the images. By analyzing the three graphs, a training dataset is automatically created and transductive learning is enabled. The transductive learner is a multi-graph based classifier, which simultaneously solves the learning problem and the problem of combining heterogeneous data. This proposed approach, overall, tackles the problem of unsupervised active learning on Web graph. Although the proposed approach is discussed in the context of WWW image retrieval, it can be applied to other domains. The experimental results show the effectiveness of our approach. Copyright © 2005 ACM.

Original language	English
Title of host publication	MIR 2005 - Proceedings of the 7th ACM SIGMM International Workshop on Multimedia Information Retrieval, Co-located with ACM Multimedia 2005
Publisher	Association for Computing Machinery
Pages	65-72
ISBN (Print)	1595932445, 9781595932440
DOIs	https://doi.org/10.1145/1101826.1101839
Publication status	Published - 10 Nov 2005
Externally published	Yes
Event	7th ACM SIGMM International Workshop on Multimedia Information Retrieval (MIR 2005) - , Singapore Duration: 10 Nov 2005 → 11 Nov 2005 https://dl.acm.org/doi/proceedings/10.1145/1101826?tocHeading=heading6

Conference

Conference	7th ACM SIGMM International Workshop on Multimedia Information Retrieval (MIR 2005)
Abbreviated title	MIR 2005
Place	Singapore
Period	10/11/05 → 11/11/05
Internet address	https://dl.acm.org/doi/proceedings/10.1145/1101826?tocHeading=heading6

Bibliographical note

Publication details (e.g. title, author(s), publication statuses and dates) are captured on an “AS IS” and “AS AVAILABLE” basis at the time of record harvesting from the data source. Suggestions for further amendments or supplementary information can be sent to [email protected].

Research Keywords

Active learning
Graph learning
Multimodal image retrieval

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access Output

10.1145/1101826.1101839

Other Access Options

Check CityUHK Library

Permanent Link

Right click to copy link

Cite this

@inproceedings{f5eb129839b24805b5252242bb801ee1,

title = "Multi-graph enabled active learning for multimodal web image retrieval",

abstract = "Tsinghua University, China In this paper, we propose a multimodal Web image retrieval technique based on multi-graph enabled active learning. The main goal is to leverage the heterogeneous data on the Web to improve retrieval precision. Three graphs are constructed on images' content features, textual annotations and hyperlinks respectively, namely Content-Graph, Text-Graph and Link-Graph, which provide complimentary information on the images. By analyzing the three graphs, a training dataset is automatically created and transductive learning is enabled. The transductive learner is a multi-graph based classifier, which simultaneously solves the learning problem and the problem of combining heterogeneous data. This proposed approach, overall, tackles the problem of unsupervised active learning on Web graph. Although the proposed approach is discussed in the context of WWW image retrieval, it can be applied to other domains. The experimental results show the effectiveness of our approach. Copyright {\textcopyright} 2005 ACM.",

keywords = "Active learning, Graph learning, Multimodal image retrieval",

author = "Xin-Jing Wang and Xing Li and Wei-Ying Ma and Lei Zhang",

note = "Publication details (e.g. title, author(s), publication statuses and dates) are captured on an “AS IS” and “AS AVAILABLE” basis at the time of record harvesting from the data source. Suggestions for further amendments or supplementary information can be sent to [email protected].; 7th ACM SIGMM International Workshop on Multimedia Information Retrieval (MIR 2005), MIR 2005 ; Conference date: 10-11-2005 Through 11-11-2005",

year = "2005",

month = nov,

day = "10",

doi = "10.1145/1101826.1101839",

language = "English",

isbn = "1595932445",

pages = "65--72",

booktitle = "MIR 2005 - Proceedings of the 7th ACM SIGMM International Workshop on Multimedia Information Retrieval, Co-located with ACM Multimedia 2005",

publisher = "Association for Computing Machinery",

address = "United States",

url = "https://dl.acm.org/doi/proceedings/10.1145/1101826?tocHeading=heading6",

}

Wang, X-J, Li, X, Ma, W-Y & Zhang, L 2005, Multi-graph enabled active learning for multimodal web image retrieval. in MIR 2005 - Proceedings of the 7th ACM SIGMM International Workshop on Multimedia Information Retrieval, Co-located with ACM Multimedia 2005. Association for Computing Machinery, pp. 65-72, 7th ACM SIGMM International Workshop on Multimedia Information Retrieval (MIR 2005), Singapore, 10/11/05. https://doi.org/10.1145/1101826.1101839

Multi-graph enabled active learning for multimodal web image retrieval. / Wang, Xin-Jing; Li, Xing; Ma, Wei-Ying et al.
MIR 2005 - Proceedings of the 7th ACM SIGMM International Workshop on Multimedia Information Retrieval, Co-located with ACM Multimedia 2005. Association for Computing Machinery, 2005. p. 65-72.

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

TY - GEN

T1 - Multi-graph enabled active learning for multimodal web image retrieval

AU - Wang, Xin-Jing

AU - Li, Xing

AU - Ma, Wei-Ying

AU - Zhang, Lei

N1 - Publication details (e.g. title, author(s), publication statuses and dates) are captured on an “AS IS” and “AS AVAILABLE” basis at the time of record harvesting from the data source. Suggestions for further amendments or supplementary information can be sent to [email protected].

PY - 2005/11/10

Y1 - 2005/11/10

N2 - Tsinghua University, China In this paper, we propose a multimodal Web image retrieval technique based on multi-graph enabled active learning. The main goal is to leverage the heterogeneous data on the Web to improve retrieval precision. Three graphs are constructed on images' content features, textual annotations and hyperlinks respectively, namely Content-Graph, Text-Graph and Link-Graph, which provide complimentary information on the images. By analyzing the three graphs, a training dataset is automatically created and transductive learning is enabled. The transductive learner is a multi-graph based classifier, which simultaneously solves the learning problem and the problem of combining heterogeneous data. This proposed approach, overall, tackles the problem of unsupervised active learning on Web graph. Although the proposed approach is discussed in the context of WWW image retrieval, it can be applied to other domains. The experimental results show the effectiveness of our approach. Copyright © 2005 ACM.

AB - Tsinghua University, China In this paper, we propose a multimodal Web image retrieval technique based on multi-graph enabled active learning. The main goal is to leverage the heterogeneous data on the Web to improve retrieval precision. Three graphs are constructed on images' content features, textual annotations and hyperlinks respectively, namely Content-Graph, Text-Graph and Link-Graph, which provide complimentary information on the images. By analyzing the three graphs, a training dataset is automatically created and transductive learning is enabled. The transductive learner is a multi-graph based classifier, which simultaneously solves the learning problem and the problem of combining heterogeneous data. This proposed approach, overall, tackles the problem of unsupervised active learning on Web graph. Although the proposed approach is discussed in the context of WWW image retrieval, it can be applied to other domains. The experimental results show the effectiveness of our approach. Copyright © 2005 ACM.

KW - Active learning

KW - Graph learning

KW - Multimodal image retrieval

UR - http://www.scopus.com/inward/record.url?scp=78650181880&partnerID=8YFLogxK

UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-78650181880&origin=recordpage

U2 - 10.1145/1101826.1101839

DO - 10.1145/1101826.1101839

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

SN - 1595932445

SN - 9781595932440

SP - 65

EP - 72

BT - MIR 2005 - Proceedings of the 7th ACM SIGMM International Workshop on Multimedia Information Retrieval, Co-located with ACM Multimedia 2005

PB - Association for Computing Machinery

T2 - 7th ACM SIGMM International Workshop on Multimedia Information Retrieval (MIR 2005)

Y2 - 10 November 2005 through 11 November 2005

ER -

Multi-graph enabled active learning for multimodal web image retrieval

Abstract

Conference

Bibliographical note

Research Keywords

UN SDGs

Access Output

Other Access Options

Permanent Link

Other Files and Links

Fingerprint

Cite this