Monitoring of epidemic outbreaks using social media data

Shinde, T.; Thatte, P.; Sachdev, S.; Pujari, V.

The COVID-19 has become the most dangerous disease for the 21st Century. The infectious disease had still gone through outbreaks despite modern medical treatments. The most recent example being the COVID-19 which has infected over 108 million people over the world and resulted in the death of over 2.3 million people as of 13 February 2021. During the ongoing pandemic of COVID-19 people are making use of social media to express their concerns as well as events related to the pandemic in their personal life. There are also a lot of agencies/organizations that are using social media platforms to convey status regarding the pandemic. We have used this overwhelming amount of data that is available on social media, particularly Twitter, to find out the trend of the COVID-19 pandemic so that we can prove the correlation between volumes of tweets tweeted related to the pandemic and daily confirmed cases which will indeed help in getting early warning regarding immediate future cases so that government and medical agencies can take appropriate measures to handle the upcoming situation. We have used natural language processing techniques and classification algorithms to classify the tweets related to the current pandemic and find the trend of the pandemic. We have used sentiment analysis techniques to find out how the current situation of the epidemic is i.e. is it getting worse or is it getting better? We have also created a live epidemic monitoring system to monitor the live tweets on a map-based UI which will show real-time tweets related to any epidemic/pandemic outbreak on a map in real-time. © 2021 IEEE.