Title: Sentiment analysis of Twitter data using machine learning: COVID-19 perspective

Authors: Shobhit Srivastava; Mrinal Kanti Sarkar; Chinmay Chakraborty

Addresses: University of Engineering & Management, Jaipur, Rajasthan 303807, India ' University of Engineering & Management, Jaipur, Rajasthan 303807, India ' BIT Mesra, Mesra, Jharkhand 835215, India

Abstract: The 2019 COVID-19 pandemic has affected people worldwide. Social media has become a global platform for individuals to voice their diverse perspectives on the pandemic, which has significantly altered their lives during and beyond lockdown periods. Twitter, a leading social media platform, experienced a surge in coronavirus-related tweets encompassing a spectrum of positive, negative and neutral opinions. Coronavirus transmits between humans in numerous ways. It irritates the lungs. This makes Twitter a perfect platform for expressing opinions. Twitter data from across the world was collected and analysed for sentiment in order to better understand public opinion and prepare for COVID-19 (Tusar et al., 2022). In this article, our aim is to compare the neural network techniques and indicate the share of their performance measures. We use kNN and neural network algorithms for these and use the MSE factor as a key of comparison. However, we use other performance measures too for better analysis of the result. Our main focus in this study is to analyse the performance partition of the kNN algorithms, including the performance portion of the each algorithm.

Keywords: COVID-19; social media; sentiment analysis; Twitter; machine learning; neural networks; KNN; neural network.

DOI: 10.1504/IJDATS.2024.137479

International Journal of Data Analysis Techniques and Strategies, 2024 Vol.16 No.1, pp.1 - 16

Received: 04 Feb 2023
Accepted: 20 Nov 2023

Published online: 19 Mar 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article