Dependency-Topic-Affects-Sentiment-LDA Model for Sentiment Analysis

Shunshun Yin, Jun Han, Yu Huang, Kuldeep Kumar

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearchpeer-review

6 Citations (Scopus)

Abstract

Sentiment analysis tends to use automated approaches to mine the sentiment information expressed in text, such as reviews, blogs and forum discussions. As most traditional approaches for sentiment analysis are based on supervised learning models and need many labeled corpora as their training data which are not always easily obtained, various unsupervised models based on Latent Dirichlet Allocation (LDA) have been proposed for sentiment classification.

In this paper, we propose a novel probabilistic modeling framework based on LDA, called Dependency-Topic-Affects-Sentiment-LDA (DTAS) model, which drops the "bag of words" assumption and assumes that the topics of sentences in a document form a Markov chain, and the sentiment of one sentence is affected by its corresponding topic and its previous sentence's topic. We applied DTAS to reviews of books and hotels. The experiment results of sentiment classification shows that DTAS outperforms other unsupervised generative models and gets high and stable accuracy.

Original languageEnglish
Title of host publicationProceedings - 2014 IEEE 26th International Conference on Tools with Artificial Intelligence, ICTAI 2014
PublisherIEEE Computer Society
Pages413-418
Number of pages6
Volume2014-December
ISBN (Electronic)9781479965724
DOIs
Publication statusPublished - 12 Dec 2014
Event26th IEEE International Conference on Tools with Artificial Intelligence, ICTAI 2014 - Limassol, Limassol, Cyprus
Duration: 10 Nov 201412 Nov 2014

Publication series

NameProceedings-International Conference on Tools With Artificial Intelligence
PublisherIEEE COMPUTER SOC
ISSN (Print)1082-3409

Conference

Conference26th IEEE International Conference on Tools with Artificial Intelligence, ICTAI 2014
CountryCyprus
CityLimassol
Period10/11/1412/11/14

Fingerprint

Blogs
Hotels
Supervised learning
Markov processes
Experiments

Cite this

Yin, S., Han, J., Huang, Y., & Kumar, K. (2014). Dependency-Topic-Affects-Sentiment-LDA Model for Sentiment Analysis. In Proceedings - 2014 IEEE 26th International Conference on Tools with Artificial Intelligence, ICTAI 2014 (Vol. 2014-December, pp. 413-418). [6984505] (Proceedings-International Conference on Tools With Artificial Intelligence). IEEE Computer Society. https://doi.org/10.1109/ICTAI.2014.69
Yin, Shunshun ; Han, Jun ; Huang, Yu ; Kumar, Kuldeep. / Dependency-Topic-Affects-Sentiment-LDA Model for Sentiment Analysis. Proceedings - 2014 IEEE 26th International Conference on Tools with Artificial Intelligence, ICTAI 2014. Vol. 2014-December IEEE Computer Society, 2014. pp. 413-418 (Proceedings-International Conference on Tools With Artificial Intelligence).
@inproceedings{59c7e3a7f52f4454927cca397a9d594b,
title = "Dependency-Topic-Affects-Sentiment-LDA Model for Sentiment Analysis",
abstract = "Sentiment analysis tends to use automated approaches to mine the sentiment information expressed in text, such as reviews, blogs and forum discussions. As most traditional approaches for sentiment analysis are based on supervised learning models and need many labeled corpora as their training data which are not always easily obtained, various unsupervised models based on Latent Dirichlet Allocation (LDA) have been proposed for sentiment classification.In this paper, we propose a novel probabilistic modeling framework based on LDA, called Dependency-Topic-Affects-Sentiment-LDA (DTAS) model, which drops the {"}bag of words{"} assumption and assumes that the topics of sentences in a document form a Markov chain, and the sentiment of one sentence is affected by its corresponding topic and its previous sentence's topic. We applied DTAS to reviews of books and hotels. The experiment results of sentiment classification shows that DTAS outperforms other unsupervised generative models and gets high and stable accuracy.",
author = "Shunshun Yin and Jun Han and Yu Huang and Kuldeep Kumar",
year = "2014",
month = "12",
day = "12",
doi = "10.1109/ICTAI.2014.69",
language = "English",
volume = "2014-December",
series = "Proceedings-International Conference on Tools With Artificial Intelligence",
publisher = "IEEE Computer Society",
pages = "413--418",
booktitle = "Proceedings - 2014 IEEE 26th International Conference on Tools with Artificial Intelligence, ICTAI 2014",
address = "United States",

}

Yin, S, Han, J, Huang, Y & Kumar, K 2014, Dependency-Topic-Affects-Sentiment-LDA Model for Sentiment Analysis. in Proceedings - 2014 IEEE 26th International Conference on Tools with Artificial Intelligence, ICTAI 2014. vol. 2014-December, 6984505, Proceedings-International Conference on Tools With Artificial Intelligence, IEEE Computer Society, pp. 413-418, 26th IEEE International Conference on Tools with Artificial Intelligence, ICTAI 2014, Limassol, Cyprus, 10/11/14. https://doi.org/10.1109/ICTAI.2014.69

Dependency-Topic-Affects-Sentiment-LDA Model for Sentiment Analysis. / Yin, Shunshun; Han, Jun; Huang, Yu; Kumar, Kuldeep.

Proceedings - 2014 IEEE 26th International Conference on Tools with Artificial Intelligence, ICTAI 2014. Vol. 2014-December IEEE Computer Society, 2014. p. 413-418 6984505 (Proceedings-International Conference on Tools With Artificial Intelligence).

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearchpeer-review

TY - GEN

T1 - Dependency-Topic-Affects-Sentiment-LDA Model for Sentiment Analysis

AU - Yin, Shunshun

AU - Han, Jun

AU - Huang, Yu

AU - Kumar, Kuldeep

PY - 2014/12/12

Y1 - 2014/12/12

N2 - Sentiment analysis tends to use automated approaches to mine the sentiment information expressed in text, such as reviews, blogs and forum discussions. As most traditional approaches for sentiment analysis are based on supervised learning models and need many labeled corpora as their training data which are not always easily obtained, various unsupervised models based on Latent Dirichlet Allocation (LDA) have been proposed for sentiment classification.In this paper, we propose a novel probabilistic modeling framework based on LDA, called Dependency-Topic-Affects-Sentiment-LDA (DTAS) model, which drops the "bag of words" assumption and assumes that the topics of sentences in a document form a Markov chain, and the sentiment of one sentence is affected by its corresponding topic and its previous sentence's topic. We applied DTAS to reviews of books and hotels. The experiment results of sentiment classification shows that DTAS outperforms other unsupervised generative models and gets high and stable accuracy.

AB - Sentiment analysis tends to use automated approaches to mine the sentiment information expressed in text, such as reviews, blogs and forum discussions. As most traditional approaches for sentiment analysis are based on supervised learning models and need many labeled corpora as their training data which are not always easily obtained, various unsupervised models based on Latent Dirichlet Allocation (LDA) have been proposed for sentiment classification.In this paper, we propose a novel probabilistic modeling framework based on LDA, called Dependency-Topic-Affects-Sentiment-LDA (DTAS) model, which drops the "bag of words" assumption and assumes that the topics of sentences in a document form a Markov chain, and the sentiment of one sentence is affected by its corresponding topic and its previous sentence's topic. We applied DTAS to reviews of books and hotels. The experiment results of sentiment classification shows that DTAS outperforms other unsupervised generative models and gets high and stable accuracy.

UR - http://www.scopus.com/inward/record.url?scp=84946541607&partnerID=8YFLogxK

U2 - 10.1109/ICTAI.2014.69

DO - 10.1109/ICTAI.2014.69

M3 - Conference contribution

VL - 2014-December

T3 - Proceedings-International Conference on Tools With Artificial Intelligence

SP - 413

EP - 418

BT - Proceedings - 2014 IEEE 26th International Conference on Tools with Artificial Intelligence, ICTAI 2014

PB - IEEE Computer Society

ER -

Yin S, Han J, Huang Y, Kumar K. Dependency-Topic-Affects-Sentiment-LDA Model for Sentiment Analysis. In Proceedings - 2014 IEEE 26th International Conference on Tools with Artificial Intelligence, ICTAI 2014. Vol. 2014-December. IEEE Computer Society. 2014. p. 413-418. 6984505. (Proceedings-International Conference on Tools With Artificial Intelligence). https://doi.org/10.1109/ICTAI.2014.69