Active learning for tag recommendation utilizing on-line photos lacking tags

Yajun Gao, Baoxin Li

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

Abstract

Recommending text tags for on-line photos is useful for Internet photo services. Typical solutions to this problem require analysis of the correlation among different attributes of the photos, including the correlation between the textual features and visual features computed from a photo. However, most on-line photos have very few tags or even no tags, and thus they contribute little or none to the analysis of tag-photo correlation, which is a key component in those schemes that rely on such analysis for tag recommendation. To address this practical challenge, we propose an active learning method for incorporating photos with no or few tags so as to enhance the correlation analysis for improved performance in tag recommendation. We demonstrate the effectiveness of the proposed approach using a dataset of more than 33,000 photos collected from Flickr.

Original languageEnglish (US)
Title of host publicationProceedings - International Conference on Image Processing, ICIP
Pages2869-2872
Number of pages4
DOIs
StatePublished - 2012
Event2012 19th IEEE International Conference on Image Processing, ICIP 2012 - Lake Buena Vista, FL, United States
Duration: Sep 30 2012Oct 3 2012

Other

Other2012 19th IEEE International Conference on Image Processing, ICIP 2012
CountryUnited States
CityLake Buena Vista, FL
Period9/30/1210/3/12

Fingerprint

Internet
Problem-Based Learning

Keywords

  • active learning
  • Tag recommendation

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Information Systems

Cite this

Gao, Y., & Li, B. (2012). Active learning for tag recommendation utilizing on-line photos lacking tags. In Proceedings - International Conference on Image Processing, ICIP (pp. 2869-2872). [6467498] https://doi.org/10.1109/ICIP.2012.6467498

Active learning for tag recommendation utilizing on-line photos lacking tags. / Gao, Yajun; Li, Baoxin.

Proceedings - International Conference on Image Processing, ICIP. 2012. p. 2869-2872 6467498.

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

Gao, Y & Li, B 2012, Active learning for tag recommendation utilizing on-line photos lacking tags. in Proceedings - International Conference on Image Processing, ICIP., 6467498, pp. 2869-2872, 2012 19th IEEE International Conference on Image Processing, ICIP 2012, Lake Buena Vista, FL, United States, 9/30/12. https://doi.org/10.1109/ICIP.2012.6467498
Gao Y, Li B. Active learning for tag recommendation utilizing on-line photos lacking tags. In Proceedings - International Conference on Image Processing, ICIP. 2012. p. 2869-2872. 6467498 https://doi.org/10.1109/ICIP.2012.6467498
Gao, Yajun ; Li, Baoxin. / Active learning for tag recommendation utilizing on-line photos lacking tags. Proceedings - International Conference on Image Processing, ICIP. 2012. pp. 2869-2872
@inproceedings{87c85aaa028e4bf5b5caaa3e3e6a0ff5,
title = "Active learning for tag recommendation utilizing on-line photos lacking tags",
abstract = "Recommending text tags for on-line photos is useful for Internet photo services. Typical solutions to this problem require analysis of the correlation among different attributes of the photos, including the correlation between the textual features and visual features computed from a photo. However, most on-line photos have very few tags or even no tags, and thus they contribute little or none to the analysis of tag-photo correlation, which is a key component in those schemes that rely on such analysis for tag recommendation. To address this practical challenge, we propose an active learning method for incorporating photos with no or few tags so as to enhance the correlation analysis for improved performance in tag recommendation. We demonstrate the effectiveness of the proposed approach using a dataset of more than 33,000 photos collected from Flickr.",
keywords = "active learning, Tag recommendation",
author = "Yajun Gao and Baoxin Li",
year = "2012",
doi = "10.1109/ICIP.2012.6467498",
language = "English (US)",
isbn = "9781467325332",
pages = "2869--2872",
booktitle = "Proceedings - International Conference on Image Processing, ICIP",

}

TY - GEN

T1 - Active learning for tag recommendation utilizing on-line photos lacking tags

AU - Gao, Yajun

AU - Li, Baoxin

PY - 2012

Y1 - 2012

N2 - Recommending text tags for on-line photos is useful for Internet photo services. Typical solutions to this problem require analysis of the correlation among different attributes of the photos, including the correlation between the textual features and visual features computed from a photo. However, most on-line photos have very few tags or even no tags, and thus they contribute little or none to the analysis of tag-photo correlation, which is a key component in those schemes that rely on such analysis for tag recommendation. To address this practical challenge, we propose an active learning method for incorporating photos with no or few tags so as to enhance the correlation analysis for improved performance in tag recommendation. We demonstrate the effectiveness of the proposed approach using a dataset of more than 33,000 photos collected from Flickr.

AB - Recommending text tags for on-line photos is useful for Internet photo services. Typical solutions to this problem require analysis of the correlation among different attributes of the photos, including the correlation between the textual features and visual features computed from a photo. However, most on-line photos have very few tags or even no tags, and thus they contribute little or none to the analysis of tag-photo correlation, which is a key component in those schemes that rely on such analysis for tag recommendation. To address this practical challenge, we propose an active learning method for incorporating photos with no or few tags so as to enhance the correlation analysis for improved performance in tag recommendation. We demonstrate the effectiveness of the proposed approach using a dataset of more than 33,000 photos collected from Flickr.

KW - active learning

KW - Tag recommendation

UR - http://www.scopus.com/inward/record.url?scp=84875857226&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84875857226&partnerID=8YFLogxK

U2 - 10.1109/ICIP.2012.6467498

DO - 10.1109/ICIP.2012.6467498

M3 - Conference contribution

AN - SCOPUS:84875857226

SN - 9781467325332

SP - 2869

EP - 2872

BT - Proceedings - International Conference on Image Processing, ICIP

ER -