Predicting the popularity of tweets by analyzing public opinion and emotions in different stages of Covid-19 pandemic

Abstract

In this study, public opinion and emotions regarding different stages of the Covid-19 pandemic from the outbreak of the disease to the distribution of vaccines were analyzed to predict the popularity of tweets. More than 1.25 million English tweets were collected, posted from January 20, 2020, to May 29, 2021. Five sets of content features, including topic analysis, topics plus TF-IDF vectorizer, bag of words (BOW) by TF-IDF vectorizer, document embedding, and document embedding plus TF-IDF vectorizer, were extracted and applied to supervised machine learning algorithms to generate a predictive model for the retweetability of posted tweets. The analysis showed that tweets with higher emotional intensity are more popular than tweets containing information on Covid-19 pandemic. This study can help to detect the public emotions during the pandemic and after vaccination and predict the retweetability of posted tweets in different stages of Covid-19 pandemic.

Social network analysis of tourism data: A case study of quarantine decisions in COVID-19 pandemic

“…Recently published studies using SNA for COVID-19 pandemic are given in Table 2 . Furthermore, several aspects of COVID-19 disease are analyzed in the literature such as prominent health issues and COVID-related public concerns using social media ( Reveilhac & Blanchard, 2022 ), public opinion and emotions regarding the COVID-19 pandemic ( Mahdikhani, 2022 ), understanding public sentiment related to COVID‐19 outbreak in Singapore ( Ridhwan & Hargreaves, 2021 ), the use of contact-tracing technology for COVID-19 disease ( Ross, 2021 ), the impact of the COVID-19 pandemic on public sector job opening ( Koch et al, 2021 ), and classification of COVID-19 cases ( Chauhan et al, 2021 ). Additionally, Karami et al (2021) used text mining to examine COVID-19 literature.…”

Exploring Pandemics Events on Twitter by Using Sentiment Analysis and Topic Modelling

“…Mahdikhani [22] combine the decision of random forest, stochastic gradient descent, and logistic regression and generate a predictive model for the retweetability of posted tweets related to COVID-19. The result shows that tweets with higher emotional intensity are more popular.…”

Self-promotion and online shaming during COVID-19: A toxic combination

“…When many people from Wuhan Hubei province in China started reporting serious health symptoms in December 2019, the coronavirus pandemic (hereafter, COVID-19) has formally begun, and since then it has spread all over the world ( Mahdikhani, 2022 ; Ridhwan and Hargreaves, 2021 ; Koch et al, 2021 ). The COVID-19 pandemic was deemed a public health emergency of international concern by the World Health Organization (WHO) in January 2020, posing a significant risk to nations with vulnerable health systems ( Karami et al, 2021 ).…”