A bayesian framework for learning shared and individual subspaces from multiple data sources

This paper presents a novel Bayesian formulation to exploit shared structures across multiple data sources, constructing foundations for effective mining and retrieval across disparate domains. We jointly analyze diverse data sources using a unifying piece of metadata (textual tags). We propose a me...

Full description

Bibliographic Details
Main Authors: Gupta, Sunil, Phung, Dinh, Adams, Brett, Venkatesh, Svetha
Other Authors: J Z Huang
Format: Conference Paper
Published: Springer 2011
Online Access:http://hdl.handle.net/20.500.11937/32333
_version_ 1848753633334657024
author Gupta, Sunil
Phung, Dinh
Adams, Brett
Venkatesh, Svetha
author2 J Z Huang
author_facet J Z Huang
Gupta, Sunil
Phung, Dinh
Adams, Brett
Venkatesh, Svetha
author_sort Gupta, Sunil
building Curtin Institutional Repository
collection Online Access
description This paper presents a novel Bayesian formulation to exploit shared structures across multiple data sources, constructing foundations for effective mining and retrieval across disparate domains. We jointly analyze diverse data sources using a unifying piece of metadata (textual tags). We propose a method based on Bayesian Probabilistic Matrix Factorization (BPMF) which is able to explicitly model the partial knowledge common to the datasets using shared subspaces and the knowledge specific to each dataset using individual subspaces. For the proposed model, we derive an efficient algorithm for learning the joint factorization based on Gibbs sampling. The effectiveness of the model is demonstrated by social media retrieval tasks across single and multiple media. The proposed solution is applicable to a wider context, providing a formal framework suitable for exploiting individual as well as mutual knowledge present across heterogeneous data sources of many kinds.
first_indexed 2025-11-14T08:27:37Z
format Conference Paper
id curtin-20.500.11937-32333
institution Curtin University Malaysia
institution_category Local University
last_indexed 2025-11-14T08:27:37Z
publishDate 2011
publisher Springer
recordtype eprints
repository_type Digital Repository
spelling curtin-20.500.11937-323332023-01-27T05:52:09Z A bayesian framework for learning shared and individual subspaces from multiple data sources Gupta, Sunil Phung, Dinh Adams, Brett Venkatesh, Svetha J Z Huang L Cao J Srivastava This paper presents a novel Bayesian formulation to exploit shared structures across multiple data sources, constructing foundations for effective mining and retrieval across disparate domains. We jointly analyze diverse data sources using a unifying piece of metadata (textual tags). We propose a method based on Bayesian Probabilistic Matrix Factorization (BPMF) which is able to explicitly model the partial knowledge common to the datasets using shared subspaces and the knowledge specific to each dataset using individual subspaces. For the proposed model, we derive an efficient algorithm for learning the joint factorization based on Gibbs sampling. The effectiveness of the model is demonstrated by social media retrieval tasks across single and multiple media. The proposed solution is applicable to a wider context, providing a formal framework suitable for exploiting individual as well as mutual knowledge present across heterogeneous data sources of many kinds. 2011 Conference Paper http://hdl.handle.net/20.500.11937/32333 10.1007/978-3-642-20841-6_12 Springer restricted
spellingShingle Gupta, Sunil
Phung, Dinh
Adams, Brett
Venkatesh, Svetha
A bayesian framework for learning shared and individual subspaces from multiple data sources
title A bayesian framework for learning shared and individual subspaces from multiple data sources
title_full A bayesian framework for learning shared and individual subspaces from multiple data sources
title_fullStr A bayesian framework for learning shared and individual subspaces from multiple data sources
title_full_unstemmed A bayesian framework for learning shared and individual subspaces from multiple data sources
title_short A bayesian framework for learning shared and individual subspaces from multiple data sources
title_sort bayesian framework for learning shared and individual subspaces from multiple data sources
url http://hdl.handle.net/20.500.11937/32333