Skip to content

matthewfarant/News-Tweets-Sentiment

Repository files navigation

Impact of Contagious Disease on Different Scale: News-Based Tweets’ Sentiments on the CBOE VIX

Abstract: The growth of social data is rapidly increasing nowadays. During the global COVID-19 outbreak, many individuals and organizations have posited their viewpoints in social media regarding the current outbreak. Twitter is widely used among the general population and is one of the main social media platforms in obtaining news and information on a topic. Currently, tweets posted by influential twitter accounts like news accounts are arguably a reliable proxy of the public sentiments’ on COVID-19. Therefore, this research intends to make good use of this information to generate insights in general, whether news twitter accounts are practicing fear mongering and whether it affects the financial market. In this research we attempt to analyze the sentiments of most-followed news accounts on twitter during COVID-19 outbreak and find out their impacts on the CBOE VIX, or also known as the “fear gauge”, one of the indicators of S&P 500 movement. Further correlation and statistical tests have found no correlation between average polarity and Chicago Board Option Exchange’s Volatility Index (CBOE VIX).

All datasets are provided in the "datasets" folder. File with "full" prefix shows full twitter dataset of an account without cleaning & filtering process. File with "clean" prefix shows cleaned & filtered Covid-19 twitter dataset. File with "nf" suffix shows full cleaned twitter dataset without filtering (Covid-19 keywords) process.

Contributors

Matthew Farant
Bernard L
Eason Cheah
Mentor: Chi Keong Lo