Twitter Sentiment Analysis in Covid-19 Pandemic

Abstract

We have yet to realise the full capability of social media as an innovative information platform during emergencies and crisis response and management. Sentiment analysis can systematically identify, extract, and scrutinise emotional states and subjective information in social media data. Exploring reactions and perceptions to response messaging is invaluable and proved especially useful for a pandemic response as it can demonstrate general population reaction to the pandemic and governments response actions. This can be further analysed to identify the gap between government response actions and communications and citizens' perceptions. In this paper, an analysis of Twitter data explores population reaction towards COVID-19 health messaging. A Natural Language Processing Python tool is known as TextBlob was used to discover general data sentiment. Data were divided into three sentiments and text extraction of health messages was conducted to explore subsequent tweets in predefined categories. Our findings show the outcome of Tweets analysis could help us to identify the general population concerns and their reactions to COVID-19 to give a better understanding of the situation to governments and support them in implementing appropriate policies.