Computational Intelligence and Healthcare Informatics

12Sentiment Analysis on Social Media for Emotional Prediction During COVID-19 Pandemic Using Efficient Machine Learning Approach

Sivanantham Kalimuthu

Research and Development Team, Crapersoft, Coimbatore

Abstract

The growth of social websites, electronic media, and blogging services is increased day by day, and they contribute many user messages like customer comments, reviews, and opinions. These mes-sages are used for measuring the quality of various products and services but it is very large. Therefore, sentiment analysis (SA) or opinion mining (OM) was introduced, and it involves in collecting the data from the web by using Natural Language Processing (NLP) techniques and examine the opinions. The opinions are collected by NLP, and then, it is classified as positive, neutral, and negative. This research investigates the SA for COVID-19 data by using various machine learning and swarm intelligence optimization algorithms. In this research, new types of SA models have been studied and performance has been evaluated for COVID-19 data by using MATLAB. The proposed HSVMCSO scheme achieves a high accuracy rate of 92.80%, sensitivity rate of 93.20%, specificity rate of 90.42%, precision rate of 95.38%, recall rate of 90.24%, and F-measure rate of 89.78% when compared with the existing SA schemes like SVM, logistic regression, and neural network. The proposed HSVMCSO attained less processing time of 29.12 s and processing cost compared to existing schemes.

Keywords: COVID-19, ...