Fast Supervised Topic Models for Short Text Emotion Detection

Jianhui PANG, Yanghui RAO, Haoran XIE, Xizhao WANG, Fu Lee WANG, Tak-lam WONG, Qing LI

Research output: Journal PublicationsJournal Article (refereed)

Abstract

With the development of social network platforms, discussion forums, and question answering websites, a huge number of short messages that typically contain a few words for an individual document are posted by online users. In these short messages, emotions are frequently embedded for communicating opinions, expressing friendship, and promoting influence. It is quite valuable to detect emotions from short messages, but the corresponding task suffers from the sparsity of feature space. In this article, we first generate term groups co-occurring in the same context to enrich the number of features. Then, two basic supervised topic models are proposed to associate emotions with topics accurately. To reduce the time cost of parameter estimation, we further propose an accelerated algorithm for our basic models. Extensive evaluations using three short corpora validate the efficiency and effectiveness of the accelerated models for predicting the emotions of unlabeled documents, in addition to generate the topic-level emotion lexicons.
Original languageEnglish
Pages (from-to)1-14
JournalIEEE Transactions on Cybernetics
DOIs
Publication statusE-pub ahead of print - 30 Sep 2019

Fingerprint

Parameter estimation
Websites
Costs

Keywords

  • Accelerated algorithm
  • emotion detection
  • short text analysis
  • topic model

Cite this

PANG, Jianhui ; RAO, Yanghui ; XIE, Haoran ; WANG, Xizhao ; WANG, Fu Lee ; WONG, Tak-lam ; LI, Qing. / Fast Supervised Topic Models for Short Text Emotion Detection. In: IEEE Transactions on Cybernetics. 2019 ; pp. 1-14.
@article{183a7f1cda33402aaa4b7c8a216c033c,
title = "Fast Supervised Topic Models for Short Text Emotion Detection",
abstract = "With the development of social network platforms, discussion forums, and question answering websites, a huge number of short messages that typically contain a few words for an individual document are posted by online users. In these short messages, emotions are frequently embedded for communicating opinions, expressing friendship, and promoting influence. It is quite valuable to detect emotions from short messages, but the corresponding task suffers from the sparsity of feature space. In this article, we first generate term groups co-occurring in the same context to enrich the number of features. Then, two basic supervised topic models are proposed to associate emotions with topics accurately. To reduce the time cost of parameter estimation, we further propose an accelerated algorithm for our basic models. Extensive evaluations using three short corpora validate the efficiency and effectiveness of the accelerated models for predicting the emotions of unlabeled documents, in addition to generate the topic-level emotion lexicons.",
keywords = "Accelerated algorithm, emotion detection, short text analysis, topic model",
author = "Jianhui PANG and Yanghui RAO and Haoran XIE and Xizhao WANG and WANG, {Fu Lee} and Tak-lam WONG and Qing LI",
year = "2019",
month = "9",
day = "30",
doi = "10.1109/TCYB.2019.2940520",
language = "English",
pages = "1--14",
journal = "IEEE Transactions on Cybernetics",
issn = "2168-2267",
publisher = "IEEE Advancing Technology for Humanity",

}

Fast Supervised Topic Models for Short Text Emotion Detection. / PANG, Jianhui; RAO, Yanghui; XIE, Haoran; WANG, Xizhao; WANG, Fu Lee; WONG, Tak-lam; LI, Qing.

In: IEEE Transactions on Cybernetics, 30.09.2019, p. 1-14.

Research output: Journal PublicationsJournal Article (refereed)

TY - JOUR

T1 - Fast Supervised Topic Models for Short Text Emotion Detection

AU - PANG, Jianhui

AU - RAO, Yanghui

AU - XIE, Haoran

AU - WANG, Xizhao

AU - WANG, Fu Lee

AU - WONG, Tak-lam

AU - LI, Qing

PY - 2019/9/30

Y1 - 2019/9/30

N2 - With the development of social network platforms, discussion forums, and question answering websites, a huge number of short messages that typically contain a few words for an individual document are posted by online users. In these short messages, emotions are frequently embedded for communicating opinions, expressing friendship, and promoting influence. It is quite valuable to detect emotions from short messages, but the corresponding task suffers from the sparsity of feature space. In this article, we first generate term groups co-occurring in the same context to enrich the number of features. Then, two basic supervised topic models are proposed to associate emotions with topics accurately. To reduce the time cost of parameter estimation, we further propose an accelerated algorithm for our basic models. Extensive evaluations using three short corpora validate the efficiency and effectiveness of the accelerated models for predicting the emotions of unlabeled documents, in addition to generate the topic-level emotion lexicons.

AB - With the development of social network platforms, discussion forums, and question answering websites, a huge number of short messages that typically contain a few words for an individual document are posted by online users. In these short messages, emotions are frequently embedded for communicating opinions, expressing friendship, and promoting influence. It is quite valuable to detect emotions from short messages, but the corresponding task suffers from the sparsity of feature space. In this article, we first generate term groups co-occurring in the same context to enrich the number of features. Then, two basic supervised topic models are proposed to associate emotions with topics accurately. To reduce the time cost of parameter estimation, we further propose an accelerated algorithm for our basic models. Extensive evaluations using three short corpora validate the efficiency and effectiveness of the accelerated models for predicting the emotions of unlabeled documents, in addition to generate the topic-level emotion lexicons.

KW - Accelerated algorithm

KW - emotion detection

KW - short text analysis

KW - topic model

U2 - 10.1109/TCYB.2019.2940520

DO - 10.1109/TCYB.2019.2940520

M3 - Journal Article (refereed)

C2 - 31567111

SP - 1

EP - 14

JO - IEEE Transactions on Cybernetics

JF - IEEE Transactions on Cybernetics

SN - 2168-2267

ER -