Classification of supervised deep learning models for COVID-19 tweets sentiment analysis
Department of Physics, Faculty of Science, University of Kufa, Najaf, Iraq.
Research Article
GSC Advanced Research and Reviews, 2024, 20(01), 408-417.
Article DOI: 10.30574/gscarr.2024.20.1.0268
Publication history:
Received on 07 June 2024; revised on 25 July 2024; accepted on 28 July 2024
Abstract:
Social media provided a successful management of human societies in light of global crises. Social media platforms have been considered the central authority in guiding society, receiving information and conducting business in many countries during the COVID-19 pandemic period in March 2020. The social platform has seen an increase in use of 45% for public platforms and 35% for the use of messages. This study suggests An AI-based model for predicting the likelihood of infection with COVID-19 through sentiment analysis and early detection using a Natural Language Processing library with deep learning techniques CNN. The model performed improved the distinction between patients who are 'positive' and patients who are 'natural' and unaffected are 'negative'. The performance of the model was tested using publicly available databases on Twitter for the period from March 16, 2020 to April 14, 2020.The achieved accuracy percentage was (~99.8% )) and based on the four measures Accuracy, Recall, Precision and F1-score.
Keywords:
Sentiment Classification; Deep Learning; Convolutional Neural Networks; COVID-19; Natural language processing
Full text article in PDF:
Copyright information:
Copyright © 2024 Author(s) retain the copyright of this article. This article is published under the terms of the Creative Commons Attribution Liscense 4.0
