Nonnegative shared subspace learning and its application to social media retrieval
Although tagging has become increasingly popular in online image and video sharing systems, tags are known to be noisy, ambiguous, incomplete and subjective. These factors can seriously affect the precision of a social tag-based web retrieval system. Therefore improving the precision performance of...
| Main Authors: | , , , , |
|---|---|
| Other Authors: | |
| Format: | Conference Paper |
| Published: |
ACM
2010
|
| Subjects: | |
| Online Access: | http://hdl.handle.net/20.500.11937/17861 |
| _version_ | 1848749580628262912 |
|---|---|
| author | Gupta, Sunil Phung, Dinh Adams, Brett Tran, Truyen Venkatesh, Svetha |
| author2 | A. Tompkins |
| author_facet | A. Tompkins Gupta, Sunil Phung, Dinh Adams, Brett Tran, Truyen Venkatesh, Svetha |
| author_sort | Gupta, Sunil |
| building | Curtin Institutional Repository |
| collection | Online Access |
| description | Although tagging has become increasingly popular in online image and video sharing systems, tags are known to be noisy, ambiguous, incomplete and subjective. These factors can seriously affect the precision of a social tag-based web retrieval system. Therefore improving the precision performance of these social tag-based web retrieval systems has become an increasingly important research topic. To this end, we propose a shared subspace learning framework to leverage a secondary source to improve retrieval performance from a primary dataset.This is achieved by learning a shared subspace between the two sources under a joint Nonnegative Matrix Factorization in which the level of subspace sharing can be explicitly controlled. We derive an efficient algorithm for learning the factorization, analyze its complexity, and provide proof of convergence. We validate the framework on image and video retrieval tasks in which tags from the LabelMe dataset are used to improve image retrieval performance from a Flickr dataset and video retrieval performance from a YouTube dataset. This has implications for how to exploit and transfer knowledge from readily available auxiliary tagging resources to improve another social web retrieval system. Our shared subspace learning framework is applicable to a range of problems where one needs to exploit the strengths existing among multiple and heterogeneous datasets. |
| first_indexed | 2025-11-14T07:23:12Z |
| format | Conference Paper |
| id | curtin-20.500.11937-17861 |
| institution | Curtin University Malaysia |
| institution_category | Local University |
| last_indexed | 2025-11-14T07:23:12Z |
| publishDate | 2010 |
| publisher | ACM |
| recordtype | eprints |
| repository_type | Digital Repository |
| spelling | curtin-20.500.11937-178612023-01-13T07:56:29Z Nonnegative shared subspace learning and its application to social media retrieval Gupta, Sunil Phung, Dinh Adams, Brett Tran, Truyen Venkatesh, Svetha A. Tompkins Q. Yang R. Bharat Rao B. Krishnapuram social media nonnegative shared subspace learning transfer learning image and video retrieval Although tagging has become increasingly popular in online image and video sharing systems, tags are known to be noisy, ambiguous, incomplete and subjective. These factors can seriously affect the precision of a social tag-based web retrieval system. Therefore improving the precision performance of these social tag-based web retrieval systems has become an increasingly important research topic. To this end, we propose a shared subspace learning framework to leverage a secondary source to improve retrieval performance from a primary dataset.This is achieved by learning a shared subspace between the two sources under a joint Nonnegative Matrix Factorization in which the level of subspace sharing can be explicitly controlled. We derive an efficient algorithm for learning the factorization, analyze its complexity, and provide proof of convergence. We validate the framework on image and video retrieval tasks in which tags from the LabelMe dataset are used to improve image retrieval performance from a Flickr dataset and video retrieval performance from a YouTube dataset. This has implications for how to exploit and transfer knowledge from readily available auxiliary tagging resources to improve another social web retrieval system. Our shared subspace learning framework is applicable to a range of problems where one needs to exploit the strengths existing among multiple and heterogeneous datasets. 2010 Conference Paper http://hdl.handle.net/20.500.11937/17861 10.1145/1835804.1835951 ACM unknown |
| spellingShingle | social media nonnegative shared subspace learning transfer learning image and video retrieval Gupta, Sunil Phung, Dinh Adams, Brett Tran, Truyen Venkatesh, Svetha Nonnegative shared subspace learning and its application to social media retrieval |
| title | Nonnegative shared subspace learning and its application to social media retrieval |
| title_full | Nonnegative shared subspace learning and its application to social media retrieval |
| title_fullStr | Nonnegative shared subspace learning and its application to social media retrieval |
| title_full_unstemmed | Nonnegative shared subspace learning and its application to social media retrieval |
| title_short | Nonnegative shared subspace learning and its application to social media retrieval |
| title_sort | nonnegative shared subspace learning and its application to social media retrieval |
| topic | social media nonnegative shared subspace learning transfer learning image and video retrieval |
| url | http://hdl.handle.net/20.500.11937/17861 |