Emotion analysis of global COVID-19 tweets using machine learning

This study presents a comprehensive analysis of global sentiment in response to the COVID-19 pandemic, focusing on tweets from various countries worldwide. Special attention was given to tweets in Arabic alongside other languages. We tested multiple preprocessing techniques and classifiers to determine the optimal combination for analyzing COVID-19-related tweets. The major emotional trends for each country were identified and visualized on a world map. Our findings reveal that anger and happiness were the predominant emotions. Notably, tweets expressing anger were most densely concentrated in South America, Russia, and the Mediterranean region, while tweets expressing happiness were most prevalent along the Indian Ocean coasts and in North America.