Weighted subspace modeling for semantic concept retrieval using gaussian mixture models

Chao Chen, Mei-Ling Shyu, Shu Ching Chen

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

At the era of digital revolution, social media data are growing at an explosive speed. Thanks to the prevailing popularity of mobile devices with cheap costs and high resolutions as well as the ubiquitous Internet access provided by mobile carriers, Wi-Fi, etc., numerous numbers of videos and pictures are generated and uploaded to social media websites such as Facebook, Flickr, and Twitter everyday. To efficiently and effectively search and retrieve information from the large amounts of multimedia data (structured, semi-structured, or unstructured), lots of algorithms and tools have been developed. Among them, a variety of data mining and machine learning methods have been explored and proposed and have shown their effectiveness and potentials in handling the growing requests to retrieve semantic information from those large-scale multimedia data. However, it is well-acknowledged that the performance of such multimedia semantic information retrieval is far from satisfactory, due to the challenges like rare events, data imbalance, etc. In this paper, a novel weighted subspace modeling framework is proposed that is based on the Gaussian Mixture Model (GMM) and is able to effectively retrieve semantic concepts, even from the highly imbalanced datasets. Experimental results performed on two public-available benchmark datasets against our previous GMM-based subspace modeling method and the other prevailing counterparts demonstrate the effectiveness of the proposed weighted GMM-based subspace modeling framework with the improved retrieval performance in terms of the mean average precision (MAP) values.

Original languageEnglish (US)
Pages (from-to)1-13
Number of pages13
JournalInformation Systems Frontiers
DOIs
StateAccepted/In press - Jun 24 2016

Fingerprint

Gaussian Mixture Model
Multimedia
Retrieval
Social Media
Semantics
Subspace
Modeling
Model-based
Semistructured Data
Subspace Methods
Wi-Fi
Rare Events
Modeling Method
Information retrieval
Mobile devices
Mobile Devices
Information Retrieval
Data mining
Learning systems
Websites

Keywords

  • Gaussian mixture model
  • Semantic concept retrieval
  • Weighted subspace modeling

ASJC Scopus subject areas

  • Information Systems
  • Computer Networks and Communications
  • Software
  • Theoretical Computer Science

Cite this

Weighted subspace modeling for semantic concept retrieval using gaussian mixture models. / Chen, Chao; Shyu, Mei-Ling; Chen, Shu Ching.

In: Information Systems Frontiers, 24.06.2016, p. 1-13.

Research output: Contribution to journalArticle

@article{6189ab50fa2c4cac99129f6777207624,
title = "Weighted subspace modeling for semantic concept retrieval using gaussian mixture models",
abstract = "At the era of digital revolution, social media data are growing at an explosive speed. Thanks to the prevailing popularity of mobile devices with cheap costs and high resolutions as well as the ubiquitous Internet access provided by mobile carriers, Wi-Fi, etc., numerous numbers of videos and pictures are generated and uploaded to social media websites such as Facebook, Flickr, and Twitter everyday. To efficiently and effectively search and retrieve information from the large amounts of multimedia data (structured, semi-structured, or unstructured), lots of algorithms and tools have been developed. Among them, a variety of data mining and machine learning methods have been explored and proposed and have shown their effectiveness and potentials in handling the growing requests to retrieve semantic information from those large-scale multimedia data. However, it is well-acknowledged that the performance of such multimedia semantic information retrieval is far from satisfactory, due to the challenges like rare events, data imbalance, etc. In this paper, a novel weighted subspace modeling framework is proposed that is based on the Gaussian Mixture Model (GMM) and is able to effectively retrieve semantic concepts, even from the highly imbalanced datasets. Experimental results performed on two public-available benchmark datasets against our previous GMM-based subspace modeling method and the other prevailing counterparts demonstrate the effectiveness of the proposed weighted GMM-based subspace modeling framework with the improved retrieval performance in terms of the mean average precision (MAP) values.",
keywords = "Gaussian mixture model, Semantic concept retrieval, Weighted subspace modeling",
author = "Chao Chen and Mei-Ling Shyu and Chen, {Shu Ching}",
year = "2016",
month = "6",
day = "24",
doi = "10.1007/s10796-016-9660-z",
language = "English (US)",
pages = "1--13",
journal = "Information Systems Frontiers",
issn = "1387-3326",
publisher = "Springer Netherlands",

}

TY - JOUR

T1 - Weighted subspace modeling for semantic concept retrieval using gaussian mixture models

AU - Chen, Chao

AU - Shyu, Mei-Ling

AU - Chen, Shu Ching

PY - 2016/6/24

Y1 - 2016/6/24

N2 - At the era of digital revolution, social media data are growing at an explosive speed. Thanks to the prevailing popularity of mobile devices with cheap costs and high resolutions as well as the ubiquitous Internet access provided by mobile carriers, Wi-Fi, etc., numerous numbers of videos and pictures are generated and uploaded to social media websites such as Facebook, Flickr, and Twitter everyday. To efficiently and effectively search and retrieve information from the large amounts of multimedia data (structured, semi-structured, or unstructured), lots of algorithms and tools have been developed. Among them, a variety of data mining and machine learning methods have been explored and proposed and have shown their effectiveness and potentials in handling the growing requests to retrieve semantic information from those large-scale multimedia data. However, it is well-acknowledged that the performance of such multimedia semantic information retrieval is far from satisfactory, due to the challenges like rare events, data imbalance, etc. In this paper, a novel weighted subspace modeling framework is proposed that is based on the Gaussian Mixture Model (GMM) and is able to effectively retrieve semantic concepts, even from the highly imbalanced datasets. Experimental results performed on two public-available benchmark datasets against our previous GMM-based subspace modeling method and the other prevailing counterparts demonstrate the effectiveness of the proposed weighted GMM-based subspace modeling framework with the improved retrieval performance in terms of the mean average precision (MAP) values.

AB - At the era of digital revolution, social media data are growing at an explosive speed. Thanks to the prevailing popularity of mobile devices with cheap costs and high resolutions as well as the ubiquitous Internet access provided by mobile carriers, Wi-Fi, etc., numerous numbers of videos and pictures are generated and uploaded to social media websites such as Facebook, Flickr, and Twitter everyday. To efficiently and effectively search and retrieve information from the large amounts of multimedia data (structured, semi-structured, or unstructured), lots of algorithms and tools have been developed. Among them, a variety of data mining and machine learning methods have been explored and proposed and have shown their effectiveness and potentials in handling the growing requests to retrieve semantic information from those large-scale multimedia data. However, it is well-acknowledged that the performance of such multimedia semantic information retrieval is far from satisfactory, due to the challenges like rare events, data imbalance, etc. In this paper, a novel weighted subspace modeling framework is proposed that is based on the Gaussian Mixture Model (GMM) and is able to effectively retrieve semantic concepts, even from the highly imbalanced datasets. Experimental results performed on two public-available benchmark datasets against our previous GMM-based subspace modeling method and the other prevailing counterparts demonstrate the effectiveness of the proposed weighted GMM-based subspace modeling framework with the improved retrieval performance in terms of the mean average precision (MAP) values.

KW - Gaussian mixture model

KW - Semantic concept retrieval

KW - Weighted subspace modeling

UR - http://www.scopus.com/inward/record.url?scp=84976294624&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84976294624&partnerID=8YFLogxK

U2 - 10.1007/s10796-016-9660-z

DO - 10.1007/s10796-016-9660-z

M3 - Article

SP - 1

EP - 13

JO - Information Systems Frontiers

JF - Information Systems Frontiers

SN - 1387-3326

ER -