Sentiment analysis: From binary to multi-class classification: A pattern-based approach for multi-class sentiment analysis in Twitter

Mondher Bouazizi, Tomoaki Ohtsuki

Research output: Chapter in Book/Report/Conference proceedingConference contribution

13 Citations (Scopus)

Abstract

Most of the state of the art works and researches on the automatic sentiment analysis and opinion mining of texts collected from social networks and microblogging websites are oriented towards the classification of texts into positive and negative. In this paper, we propose a pattern-based approach that goes deeper in the classification of texts collected from Twitter (i.e., tweets). We classify the tweets into 7 different classes; however the approach can be run to classify into more classes. Experiments show that our approach reaches an accuracy of classification equal to 56.9% and a precision level of sentimental tweets (other than neutral and sarcastic) equal to 72.58%. Nevertheless, the approach proves to be very accurate in binary classification (i.e., classification into positive and negative) and ternary classification (i.e., classification into positive, negative and neutral): in the former case, we reach an accuracy of 87.5% for the same dataset used after removing neutral tweets, and in the latter case, we reached an accuracy of classification of 83.0%.

Original languageEnglish
Title of host publication2016 IEEE International Conference on Communications, ICC 2016
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781479966646
DOIs
Publication statusPublished - 2016 Jul 12
Event2016 IEEE International Conference on Communications, ICC 2016 - Kuala Lumpur, Malaysia
Duration: 2016 May 222016 May 27

Other

Other2016 IEEE International Conference on Communications, ICC 2016
CountryMalaysia
CityKuala Lumpur
Period16/5/2216/5/27

Fingerprint

Websites
Experiments

ASJC Scopus subject areas

  • Computer Networks and Communications

Cite this

Bouazizi, M., & Ohtsuki, T. (2016). Sentiment analysis: From binary to multi-class classification: A pattern-based approach for multi-class sentiment analysis in Twitter. In 2016 IEEE International Conference on Communications, ICC 2016 [7511392] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICC.2016.7511392

Sentiment analysis : From binary to multi-class classification: A pattern-based approach for multi-class sentiment analysis in Twitter. / Bouazizi, Mondher; Ohtsuki, Tomoaki.

2016 IEEE International Conference on Communications, ICC 2016. Institute of Electrical and Electronics Engineers Inc., 2016. 7511392.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Bouazizi, M & Ohtsuki, T 2016, Sentiment analysis: From binary to multi-class classification: A pattern-based approach for multi-class sentiment analysis in Twitter. in 2016 IEEE International Conference on Communications, ICC 2016., 7511392, Institute of Electrical and Electronics Engineers Inc., 2016 IEEE International Conference on Communications, ICC 2016, Kuala Lumpur, Malaysia, 16/5/22. https://doi.org/10.1109/ICC.2016.7511392
Bouazizi M, Ohtsuki T. Sentiment analysis: From binary to multi-class classification: A pattern-based approach for multi-class sentiment analysis in Twitter. In 2016 IEEE International Conference on Communications, ICC 2016. Institute of Electrical and Electronics Engineers Inc. 2016. 7511392 https://doi.org/10.1109/ICC.2016.7511392
Bouazizi, Mondher ; Ohtsuki, Tomoaki. / Sentiment analysis : From binary to multi-class classification: A pattern-based approach for multi-class sentiment analysis in Twitter. 2016 IEEE International Conference on Communications, ICC 2016. Institute of Electrical and Electronics Engineers Inc., 2016.
@inproceedings{d002542093ef491b8527f12f9f5b495a,
title = "Sentiment analysis: From binary to multi-class classification: A pattern-based approach for multi-class sentiment analysis in Twitter",
abstract = "Most of the state of the art works and researches on the automatic sentiment analysis and opinion mining of texts collected from social networks and microblogging websites are oriented towards the classification of texts into positive and negative. In this paper, we propose a pattern-based approach that goes deeper in the classification of texts collected from Twitter (i.e., tweets). We classify the tweets into 7 different classes; however the approach can be run to classify into more classes. Experiments show that our approach reaches an accuracy of classification equal to 56.9{\%} and a precision level of sentimental tweets (other than neutral and sarcastic) equal to 72.58{\%}. Nevertheless, the approach proves to be very accurate in binary classification (i.e., classification into positive and negative) and ternary classification (i.e., classification into positive, negative and neutral): in the former case, we reach an accuracy of 87.5{\%} for the same dataset used after removing neutral tweets, and in the latter case, we reached an accuracy of classification of 83.0{\%}.",
author = "Mondher Bouazizi and Tomoaki Ohtsuki",
year = "2016",
month = "7",
day = "12",
doi = "10.1109/ICC.2016.7511392",
language = "English",
booktitle = "2016 IEEE International Conference on Communications, ICC 2016",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - Sentiment analysis

T2 - From binary to multi-class classification: A pattern-based approach for multi-class sentiment analysis in Twitter

AU - Bouazizi, Mondher

AU - Ohtsuki, Tomoaki

PY - 2016/7/12

Y1 - 2016/7/12

N2 - Most of the state of the art works and researches on the automatic sentiment analysis and opinion mining of texts collected from social networks and microblogging websites are oriented towards the classification of texts into positive and negative. In this paper, we propose a pattern-based approach that goes deeper in the classification of texts collected from Twitter (i.e., tweets). We classify the tweets into 7 different classes; however the approach can be run to classify into more classes. Experiments show that our approach reaches an accuracy of classification equal to 56.9% and a precision level of sentimental tweets (other than neutral and sarcastic) equal to 72.58%. Nevertheless, the approach proves to be very accurate in binary classification (i.e., classification into positive and negative) and ternary classification (i.e., classification into positive, negative and neutral): in the former case, we reach an accuracy of 87.5% for the same dataset used after removing neutral tweets, and in the latter case, we reached an accuracy of classification of 83.0%.

AB - Most of the state of the art works and researches on the automatic sentiment analysis and opinion mining of texts collected from social networks and microblogging websites are oriented towards the classification of texts into positive and negative. In this paper, we propose a pattern-based approach that goes deeper in the classification of texts collected from Twitter (i.e., tweets). We classify the tweets into 7 different classes; however the approach can be run to classify into more classes. Experiments show that our approach reaches an accuracy of classification equal to 56.9% and a precision level of sentimental tweets (other than neutral and sarcastic) equal to 72.58%. Nevertheless, the approach proves to be very accurate in binary classification (i.e., classification into positive and negative) and ternary classification (i.e., classification into positive, negative and neutral): in the former case, we reach an accuracy of 87.5% for the same dataset used after removing neutral tweets, and in the latter case, we reached an accuracy of classification of 83.0%.

UR - http://www.scopus.com/inward/record.url?scp=84981295687&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84981295687&partnerID=8YFLogxK

U2 - 10.1109/ICC.2016.7511392

DO - 10.1109/ICC.2016.7511392

M3 - Conference contribution

AN - SCOPUS:84981295687

BT - 2016 IEEE International Conference on Communications, ICC 2016

PB - Institute of Electrical and Electronics Engineers Inc.

ER -