Sentiment Analysis of Social Media Comments in Mauritius

Sahib, Nuzhah Gooda; Marianne, Marie Angele Justine; Gobin-Rahimbux, Baby

doi:10.1109/ccwc57344.2023.10099291

Cited by 1 publication

(1 citation statement)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The MNB method was applied by Rahman et al [32] to classify sentiment in Bengali films, and the technique gave an accuracy of 86%. Several other studies on sentiment classification using the MNB method [7,10,32,35,38].…”

Section: Introductionmentioning

confidence: 99%

Unbalanced Multiclass Classification With Adaptive Synthetic Multinomial Naive Bayes Approach

Fauzi,

Ismatullah,

Manfaati Nur

2023

IAPGOS

View full text Add to dashboard Cite

Opinions related to rising fuel prices need to be seen and analysed. Public opinion is closely related to public policy in Indonesia in the future. Twitter is one of the media that people use to convey their opinions. This study uses sentiment analysis to look at this phenomenon. Sentiment is divided into three categories: positive, neutral, and negative. The methods used in this research are Adaptive Synthetic Multinomial Naive Bayes, Adaptive Synthetic k-nearest neighbours, and Adaptive Synthetic Random Forest. The Adaptive Synthetic method is used to handle unbalanced data. The data used in this study are public arguments per province in Indonesia. The results obtained in this study are negative sentiments that dominate all provinces in Indonesia. There is a relationship between negative sentiment and the level of education, internet use, and the human development index. Adaptive Synthetic Multinomial Naive Bayes performed better than other methods, with an accuracy of 0.882. The highest accuracy of the Adaptive Synthetic Multinomial Naive Bayes method is 0.990 in Papua Barat Province.

show abstract