COVID-19 Population sentiment analysis

This study explores the reaction of different cultures to mandatory mask-wearing measures.

Introduction

Since the declaration of COVID-19 as a pandemic, responses to mask-wearing mandates have varied across countries. This project investigates these reactions by analyzing Twitter data, focusing on the differences between collectivistic, moderate, and individualistic cultures.

Methodology

Data Collection

Tweets containing the keyword "mask" were collected using Twint, a Twitter scraping tool.

Data Pre-processing

Tweets were cleaned using Pandas and the NeatText library for NLP.

Data Analysis

Sentiment analysis was performed with CoreNLP, followed by topic modeling using Mallet to explore the main topics within each sentiment category.

Results

Sentiment Analysis

Negative sentiments were prevalent across all countries, with specific nuances based on cultural context and current events.

Topic Modeling

Topic modeling revealed common concerns and topics among tweets with similar sentiments, providing insights into the specific issues and discussions surrounding mask-wearing in each country.

Conclusion

The study highlights the complexity of public sentiment towards mask-wearing mandates, influenced by cultural differences and individual perspectives.