Sentiment Analysis of Social Media Response and Spatial Distribution Patterns on the COVID-19 Outbreak: The Case Study of Italy

The study was carried out on people’s exchange on social media networks, particularly on Twitter, during the COVID-19 pandemic. We analyzed over 4 million COVID-19 related tweets classified as fear, anger, and joy in four of Italy’s geographic regions to investigate whether socioeconomic factors and sentiments of tweets shift over the course of the pandemic and when lagged to specific policy shifts before and after the lock-down. The result shows that the north of Italy was found to have the highest number of social and environmental impacts when compared to the other regions. However, southern Italy and islands were most hit by the economic crisis, including the shortage of medical supplies, weaker infrastructure, loss of tourism, and high unemployment rate. Thus, the pandemic widened the existing gap between the north and the south regions of Italy, emphasizing the lack of preparedness and delays implemented by governments when addressing the emergency, lack of response to a nationwide emergency and communication plan, and decentralization of the health care system. By tracking and analyzing health communications and their impacts, this study demonstrates value of social media and geospatial sentiment analysis in supporting decision making for policy makers or health care professionals.