Home / Current Issue / Paper 1705468
BERT-Based Sentiment Analysis of Indian COVID-19 Tweets for Policy Making
Subject area: Science,Engineering and Technology · Area of research: NLP, Machine Learning
Abstract
In the unprecedented global crisis of COVID-19, social media platforms have emerged as vital communication channels, reflecting public sentiment and opinion. This paper presents an in-depth analysis of Indian Twitter data during the COVID-19 pandemic, employing the Bidirectional Encoder Representations from Transformers (BERT) model for sentiment analysis. The primary objective is to explore the public sentiment in India regarding the pandemic and its implications for policy-making. The study systematically collected a significant corpus of tweets related to COVID-19 in India, ensuring a diverse representation of opinions and perspectives. Following rigorous data pre-processing, including cleaning and tokenization, the BERT model was fine-tuned to suit the linguistic nuances and contextual intricacies of the tweets. The sentiment analysis focused on categorizing tweets into various emotional responses, such as fear, anger, optimism, and support, providing a nuanced understanding of the public's perception. The results yielded by the BERT-based analysis offer insightful and nuanced categorizations of sentiments, which are critical in informing and guiding policy decisions. These findings not only underscore the potential of advanced NLP techniques in public sentiment analysis but also highlight their practical implications in real-time policy-making, especially during crisis situations. The study contributes to the growing body of literature on sentiment analysis using deep learning and offers a novel perspective on leveraging social media data for governmental decision-making in health emergencies.
Keywords
BERT, COVID-19, Data Analytics, India, Machine Learning, Natural Language Processing, Policy Making, Public Sentiment, Sentiment Analysis, Transformer Model, Twitter Data.
How to cite this paper
@article{1705468,
author = {Pranav Khot, Mithilesh Vishwakarma, Vidhi Shukla},
title = {BERT-Based Sentiment Analysis of Indian COVID-19 Tweets for Policy Making},
journal = {Iconic Research And Engineering Journals},
year = {2024},
volume = {7},
number = {8},
pages = {26-33},
issn = {2456-8880},
url = {https://www.irejournals.com/formatedpaper/1705468.pdf},
abstract = {In the unprecedented global crisis of COVID-19, social media platforms have emerged as vital communication channels, reflecting public sentiment and opinion. This paper presents an in-depth analysis of Indian Twitter data during the COVID-19 pandemic, employing the Bidirectional Encoder Representations from Transformers (BERT) model for sentiment analysis. The primary objective is to explore the public sentiment in India regarding the pandemic and its implications for policy-making. The study systematically collected a significant corpus of tweets related to COVID-19 in India, ensuring a diverse representation of opinions and perspectives. Following rigorous data pre-processing, including cleaning and tokenization, the BERT model was fine-tuned to suit the linguistic nuances and contextual intricacies of the tweets. The sentiment analysis focused on categorizing tweets into various emotional responses, such as fear, anger, optimism, and support, providing a nuanced understanding of the public's perception. The results yielded by the BERT-based analysis offer insightful and nuanced categorizations of sentiments, which are critical in informing and guiding policy decisions. These findings not only underscore the potential of advanced NLP techniques in public sentiment analysis but also highlight their practical implications in real-time policy-making, especially during crisis situations. The study contributes to the growing body of literature on sentiment analysis using deep learning and offers a novel perspective on leveraging social media data for governmental decision-making in health emergencies.},
keywords = {BERT, COVID-19, Data Analytics, India, Machine Learning, Natural Language Processing, Policy Making, Public Sentiment, Sentiment Analysis, Transformer Model, Twitter Data.},
month = {February},
}