 Sentiment analysis and topic modeling of COVID-19 tweets of IndiaManju Bhardwaj (),
Priya Mishra (),
Shikha Badhani () and
Sunil K. Muttoo ()
International Journal of System Assurance Engineering and Management, 2024, vol. 15, issue 5, No 12, 1756-1776 Abstract: Abstract Social media platforms provide an opportunity to the users to express their views and emotions on any topic. Various researchers have successfully used the content posted on these platforms to capture the emotions of the people about the given event or topic. During COVID-19 pandemic, Indians extensively used Twitter owing to an increased need for virtual interaction. In this work, we analyse the tweets posted in India during COVID-19 outbreak to understand how individuals in India reacted to the pandemic. We identified the timelines of three major COVID-19 waves from May 2020 to March 2022 and retrieved 13,818 tweets from COV19Tweets dataset available at IEEE DataPort for the respective duration of each of the three waves. Lexicon based sentiment analysis of the tweets indicated a positive mindset of the Indian population during the pandemic. Further, visual analysis through word clouds revealed that a few words were common for all waves whereas some words were wave-specific. It was observed that the words used in tweets cannot be compulsorily associated with positive or negative emotions, as the context or the set of words taken together may be a better indicator. Hence, machine learning approach was followed for the identification of sentiments by extracting BoW (Bag-of-Words) and TF–IDF (Term Frequency–Inverse Document Frequency) features from the tweet text. Comparative performance analysis of the four classification algorithms, namely, Decision Tree (DT), Logistic Regression (LR), Naive Bayes (NB), and Support Vector Machines (SVM) and two ensemble methods Adaboost and Random Forest revealed that LR applied to BoW featureset was the best performer. Finally, we performed Latent Dirichlet Allocation (LDA) based topic modeling on the COVID-19 tweets to identify topics of discussion in each of the waves. The topics evolved from informative messages related to the pandemic during the first wave, to wider discussions related to the impact of COVID-19 on nifty, tourism, etc. for the second wave, and the omicron virus, availability of beds, and ventilators in the third wave. This study can be of great interest to governments, as they may undertake similar studies to understand human behavior when natural calamities or pandemics occur at the local or global levels. The automated capture of public sentiments and identification of topics may expedite the appropriate execution of preventive measures taken by governments and address the concerns of citizens almost instantly. Keywords: COVID-19; Machine learning; Natural Language Processing (NLP); Sentiment analysis; Twitter data; Visualization; Topic modeling; Latent Dirichlet Allocation; Lexicon; Bag-Of-Words (search for similar items in EconPapers) Downloads: (external link) Related works: Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text Persistent link: https://EconPapers.repec.org/RePEc:spr:ijsaem:v:15:y:2024:i:5:d:10.1007_s13198-023-02082-0 Ordering information: This journal article can be ordered from DOI: 10.1007/s13198-023-02082-0 Access Statistics for this article International Journal of System Assurance Engineering and Management is currently edited by P.K. Kapur, A.K. Verma and U. Kumar More articles in International Journal of System Assurance Engineering and Management from Springer, The Society for Reliability, Engineering Quality and Operations Management (SREQOM),India, and Division of Operation and Maintenance, Lulea University of Technology, Sweden |
Is your work missing from RePEc? Here is how to contribute.
Questions or problems? Check the EconPapers FAQ or send mail to .
EconPapers is hosted by the
School of Business at Örebro University.