COVID-19 Public Sentiment Insights and Machine Learning for Tweets Classification

• DOI: 10.20944/preprints202005.0015.v1
• Publication Date: 2020-05-04T16:19:38Z
• Authors: Jim Samuel; G. G. Md. Nawaz Ali; Md. Mokhlesur Rahman; Ek Esawi; Yana Samuel
• Keywords: Sentiment Analysis; Social media analytics

Along with the Coronavirus pandemic, another crisis has manifested itself in form of mass fear and panic phenomena, fuelled by incomplete often inaccurate information. There is therefore a tremendous need to address better understand COVID-19's informational gauge public sentiment, so that appropriate messaging policy decisions can be implemented. In this research article, we identify sentiment associated pandemic using specific Tweets R statistical software, along its analysis packages. We demonstrate insights into progress fear-sentiment over time as COVID-19 approached peak levels United States, descriptive textual analytics supported necessary data visualizations. Furthermore, provide methodological overview two essential machine learning classification methods, context analytics, compare their effectiveness classifying varying lengths. observe strong accuracy 91% for short Tweets, Naive Bayes method. also logistic regression method provides reasonable 74% shorter both methods showed relatively weaker performance longer Tweets. This progression, outlines implications, limitations opportunities.