COVID-19: A Study on Emotions

सार

COVID-19 has created an unprecedented crisis. With the number of coronavirus (COVID-19) cases growing exponentially, the entire world has come to a standstill. This COVID-19 global pandemic has created stressful unsettling circumstances for the whole world. It has completely disrupted the normal socio-economic activities. People have flooded their social media accounts with their views and comments in these circumstances. In this paper, we have extracted tweets for two months on a weekly basis. A total of nine lac tweets are extracted in the course of two months. The extracted tweets are pre-processed and a dataset is made for the month of March and April 2020. The research is based on emotional analysis classifier using computer software programmes count vectoriser, tfidf transformer and logistic regression. Also, human emotions are classified into four broad categories: Anger, Fear, Joy, Sadness. The model is trained on a training dataset and tested on our dataset to retrieve the emotions of tweets. The results prove that people have a lot of fear in their mind and they are worried due to COVID19. The results also depict that there is a geographical shift of density of fear emotion from March to April. This is because of the shift in epicenter of coronavirus from Europe to USA. The paper has also listed some suggestions to improve the emotion and mental health of the people as well as to avoid the panic and stress created due to COVID-19 global pandemic.