Indian Journal of Science and Technology
Year: 2023, Volume: 16, Issue: 47, Pages: 4561-4568
Original Article
R B Alemayehu1, M Rout2*, S C Satapathy3
1Research Scholar, School of CE, KIIT Deemed to be University, Bhubaneswar, 751 024, Odisha, India
2Associate Professor, School of CE, KIIT Deemed to be University, Bhubaneswar, 751 024, Odisha, India
3Professor & Dean (Research), KIIT Deemed to be University, Bhubaneswar, 751 024, Odisha, India
*Corresponding Author
Email: [email protected]
Received Date:15 March 2023, Accepted Date:19 July 2023, Published Date:30 December 2023
Objectives: This study aims to prepare a corpus and explore sentiment analysis in the Amharic language, which is increasingly used due to the growth of both the language and the Internet. Methods: The study acquired 23,646 Amharic tweets from Twitter using the Twitter API, cleaned and normalized the text through preprocessing, and manually annotated the data as positive, negative, or neutral by three annotators. The study utilized a multi-scale sentiment analysis approach to experimentally evaluate the classifier's performance and compare different ML and DL classifiers. Findings: The study found that sentiment analysis in the Amharic language in this dataset showed that the KNN classifier could classify texts with an accuracy of 76% and 90% accuracy using the CNN deep learning classifier. Novelty: This study contributes to the field of sentiment analysis by addressing the scarcity of an Amharic-language dataset specifically tailored for sentiment analysis purposes. Our approach involves filling this critical research gap by developing a new dataset. Subsequently, we employ machine learning and deep learning classifiers to assess the viability of this dataset for performing multi-class sentiment analysis tasks in the Amharic language.
Keywords: Amharic, Sentiment Analysis, Multiclass, Machine Learning, Deep Learning Classifier
© 2023 Alemayehu et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Published By Indian Society for Education and Environment (iSee)
Subscribe now for latest articles and news.