ANALYSIS OF COVID-19 CASES AND PEOPLE EMOTIONS USING MACHINE LEARNING

Abstract

In recent days the entire world is facing and fighting towards a severe health problem that is known as COVID-19. Earlier COVID-19 is known as novel coronavirus. As of now, there are a total of 9,711,197 cases that have been reported and found positive in the entire world among more than 200 countries. The number of deaths reported are a huge in number. Four lakhs ninety-one thousand and seven hundred ninety-three (4,91,793). The total number of cases that have been recovered are 5,247,173 in numbers. Doing analysis on such a vast dataset is itself an exciting and challenging task. In recent days and from the beginning of the year 2020 sentiment and emotions of people have also changed and developed in a various manner due to events happening around in their environment and surroundings. People's sentiment, emotions, and opinions are a beneficial medium for analysing the recent trend in society. People's views emotions and ideas also convey knowledge and information about how people are reacting to a particular event in the community, states, cities, countries, and the world. One such event is COVID-19, COVID-19 is also known as Coronavirus Disease. These opinions shared by people can vary in many forms, such as videos, images, podcasts, audio, and text. People generally share their emotions on social media platforms those are widely used now a days and are Facebook, Instagram, Twitter, YouTube, Blogs, etc. In this research work, we have focused on text data extracted by twitter using twitter API during the period of COVID-19. We will be finding sentimental analysis of twitter users during the period of COVID-19. With data that is vast and v significant, which is extracted from twitter using various hashtags related to coronavirus, COVID-19, China, Italy, Trump, etc. We will be performing an analysis of positive as well as negative sentiment displayed in people, various user tweets around the globe marked with different hashtags that more likely to be delivered via the negative emotion. Notably, we introduce a stage-based method to entertain wherewith the negative sentiment changes simultaneously with unique various development frames of COVID19, which changed from a society residential community epidemic into the domestic level and a worldwide public health emergency furthermore later, into the global pandemic. At each stage of COVID-19 Coronavirus, sentiment analysis allows us to understand the sentiment from tweets that can be majorly negative in essence with various hashtags. Furthermore, the extraction of keywords renders for the development of ideas in the definition of negative emotion through specific tweets.