CS7DS4 - Trinity College Dublin, MSc Computer Science - Data Science, 2020
1 Introduction
Twitter is an integral part of contemporary social life. It has a 330 million active users’ database, with about 500 million tweets sent in a single day [1]. This makes it play a vital role in trend discovery and thought sharing in a unique and engaging way. With the recent outbreak of the Covid-19 pandemic, and many countries announcing strict lockdowns and curfews, twitter saw an unprecedented surge in tweets. As of 14th April, there have been 508 million tweets with the hashtags #covid19 and #coronavirus alone [2].
Moreover, even the World Health Organization has been using Twitter as a platform to engage with people and advising people about the pandemic. W.H.O. has tweeted more than 1800 times since 1st January, compared to about 600 tweets last year in equivalent duration [2]. It is evident that Twitter is fundamental when it comes to connecting with people across the world.
2 Motivation
On 22nd March, the government of India initiated a complete lockdown as a response to the growing number of Covid-19 cases in India. Besides, many other countries were already in some form and state of a lockdown during this period. Perceptibly, people everywhere took to social media to express their support and complaints about these initiatives.
It would be interesting to analyze and visualize in changing trends of what people tweeted during this lockdown. It would also be interesting to see how each Country/State/City is dealing with the pandemic and their general views about the same. Following visualization could be used:
1) Trends in the most frequent Terms or single Grams occurring in tweets.
2) Trends in the most frequent Bigrams or combination of two grams occurring in tweets.
3) Trends in the most frequent Trigrams or combination of three grams occurring in tweets.
4) A general sentiment analysis performed on the tweets, categorized by Country/State/City.
These trends would reveal a lot about tweeting habits of a country, and users in general. No visualization, the best of my knowledge, achieves this goal.
References
[1] “Twitter for Business,” Twitter, Inc., [Online]. Available: [ Ссылка ]. [Accessed 16 April 2020].
[2] “Covid-19 - Twitter Evolution of Coronavirus - Tweet Binder,” Tweet Binder, 2020. [Online]. Available: [ Ссылка ]. [Accessed 16 April 2020].
