Affect analysis of web forums and blogs using correlation ensembles

Ahmed Abbasi, Hsinchun Chen, Sven Thoms, Tianjun Fu

Research output: Contribution to journalArticle

91 Citations (Scopus)

Abstract

Analysis of affective intensities in computer-mediated communication is important in order to allow a better understanding of online users' emotions and preferences. Despite considerable research on textual affect classification, it is unclear which features and techniques are most effective. In this study, we compared several feature representations for affect analysis, including learned n-grams and various automatically and manually crafted affect lexicons. We also proposed the support vector regression correlation ensemble (SVRCE) method for enhanced classification of affect intensities. SVRCE uses an ensemble of classifiers each trained using a feature subset tailored toward classifying a single affect class. The ensemble is combined with affect correlation information to enable better prediction of emotive intensities. Experiments were conducted on four test beds encompassing web forums, blogs, and online stories. The results revealed that learned n-grams were more effective than lexicon-based affect representations. The findings also indicated that SVRCE outperformed comparison techniques, including Pace regression, semantic orientation, and WordNet models. Ablation testing showed that the improved performance of SVRCE was attributable to its use of feature ensembles as well as affect correlation information. A brief case study was conducted to illustrate the utility of the features and techniques for affect analysis of large archives of online discourse.

Original languageEnglish (US)
Article number4479460
Pages (from-to)1168-1180
Number of pages13
JournalIEEE Transactions on Knowledge and Data Engineering
Volume20
Issue number9
DOIs
StatePublished - Sep 2008

Fingerprint

Blogs
Correlation methods
Ablation
Classifiers
Semantics
Communication
Testing
Experiments

Keywords

  • Affective computing
  • Discourse
  • Emotion recognition
  • Linguistic processing
  • Machine learning
  • Text mining

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Electrical and Electronic Engineering
  • Artificial Intelligence
  • Information Systems

Cite this

Affect analysis of web forums and blogs using correlation ensembles. / Abbasi, Ahmed; Chen, Hsinchun; Thoms, Sven; Fu, Tianjun.

In: IEEE Transactions on Knowledge and Data Engineering, Vol. 20, No. 9, 4479460, 09.2008, p. 1168-1180.

Research output: Contribution to journalArticle

Abbasi, Ahmed ; Chen, Hsinchun ; Thoms, Sven ; Fu, Tianjun. / Affect analysis of web forums and blogs using correlation ensembles. In: IEEE Transactions on Knowledge and Data Engineering. 2008 ; Vol. 20, No. 9. pp. 1168-1180.
@article{de47c318937f4735be1d614b7117b4b9,
title = "Affect analysis of web forums and blogs using correlation ensembles",
abstract = "Analysis of affective intensities in computer-mediated communication is important in order to allow a better understanding of online users' emotions and preferences. Despite considerable research on textual affect classification, it is unclear which features and techniques are most effective. In this study, we compared several feature representations for affect analysis, including learned n-grams and various automatically and manually crafted affect lexicons. We also proposed the support vector regression correlation ensemble (SVRCE) method for enhanced classification of affect intensities. SVRCE uses an ensemble of classifiers each trained using a feature subset tailored toward classifying a single affect class. The ensemble is combined with affect correlation information to enable better prediction of emotive intensities. Experiments were conducted on four test beds encompassing web forums, blogs, and online stories. The results revealed that learned n-grams were more effective than lexicon-based affect representations. The findings also indicated that SVRCE outperformed comparison techniques, including Pace regression, semantic orientation, and WordNet models. Ablation testing showed that the improved performance of SVRCE was attributable to its use of feature ensembles as well as affect correlation information. A brief case study was conducted to illustrate the utility of the features and techniques for affect analysis of large archives of online discourse.",
keywords = "Affective computing, Discourse, Emotion recognition, Linguistic processing, Machine learning, Text mining",
author = "Ahmed Abbasi and Hsinchun Chen and Sven Thoms and Tianjun Fu",
year = "2008",
month = "9",
doi = "10.1109/TKDE.2008.51",
language = "English (US)",
volume = "20",
pages = "1168--1180",
journal = "IEEE Transactions on Knowledge and Data Engineering",
issn = "1041-4347",
publisher = "IEEE Computer Society",
number = "9",

}

TY - JOUR

T1 - Affect analysis of web forums and blogs using correlation ensembles

AU - Abbasi, Ahmed

AU - Chen, Hsinchun

AU - Thoms, Sven

AU - Fu, Tianjun

PY - 2008/9

Y1 - 2008/9

N2 - Analysis of affective intensities in computer-mediated communication is important in order to allow a better understanding of online users' emotions and preferences. Despite considerable research on textual affect classification, it is unclear which features and techniques are most effective. In this study, we compared several feature representations for affect analysis, including learned n-grams and various automatically and manually crafted affect lexicons. We also proposed the support vector regression correlation ensemble (SVRCE) method for enhanced classification of affect intensities. SVRCE uses an ensemble of classifiers each trained using a feature subset tailored toward classifying a single affect class. The ensemble is combined with affect correlation information to enable better prediction of emotive intensities. Experiments were conducted on four test beds encompassing web forums, blogs, and online stories. The results revealed that learned n-grams were more effective than lexicon-based affect representations. The findings also indicated that SVRCE outperformed comparison techniques, including Pace regression, semantic orientation, and WordNet models. Ablation testing showed that the improved performance of SVRCE was attributable to its use of feature ensembles as well as affect correlation information. A brief case study was conducted to illustrate the utility of the features and techniques for affect analysis of large archives of online discourse.

AB - Analysis of affective intensities in computer-mediated communication is important in order to allow a better understanding of online users' emotions and preferences. Despite considerable research on textual affect classification, it is unclear which features and techniques are most effective. In this study, we compared several feature representations for affect analysis, including learned n-grams and various automatically and manually crafted affect lexicons. We also proposed the support vector regression correlation ensemble (SVRCE) method for enhanced classification of affect intensities. SVRCE uses an ensemble of classifiers each trained using a feature subset tailored toward classifying a single affect class. The ensemble is combined with affect correlation information to enable better prediction of emotive intensities. Experiments were conducted on four test beds encompassing web forums, blogs, and online stories. The results revealed that learned n-grams were more effective than lexicon-based affect representations. The findings also indicated that SVRCE outperformed comparison techniques, including Pace regression, semantic orientation, and WordNet models. Ablation testing showed that the improved performance of SVRCE was attributable to its use of feature ensembles as well as affect correlation information. A brief case study was conducted to illustrate the utility of the features and techniques for affect analysis of large archives of online discourse.

KW - Affective computing

KW - Discourse

KW - Emotion recognition

KW - Linguistic processing

KW - Machine learning

KW - Text mining

UR - http://www.scopus.com/inward/record.url?scp=50649122516&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=50649122516&partnerID=8YFLogxK

U2 - 10.1109/TKDE.2008.51

DO - 10.1109/TKDE.2008.51

M3 - Article

AN - SCOPUS:50649122516

VL - 20

SP - 1168

EP - 1180

JO - IEEE Transactions on Knowledge and Data Engineering

JF - IEEE Transactions on Knowledge and Data Engineering

SN - 1041-4347

IS - 9

M1 - 4479460

ER -