USING SENTIMENT AND EMOTION ANALYSIS OF NEWS ARTICLES TO ANALYZE THE EFFECTS OF LEADER’S STATEMENTS ON COVID-19 SPREAD

Abstract

Leaders generally include government officials, politicians, etc. Their statements can highly affect people’s decisions in many ways. Currently, in the pandemic situation, many statements were being passed every hour and day, which showed an impact on the spread of corona virus cases at certain location. So, this paper proposes a supervised model to analyze the variations of COVID-19 data based upon the leader’s statements passed at certain time and location. The proposed methodology consists of sentiment and emotion analysis for the leader’s statements to determine the true intentions of the leader. The leader’s statements are a collection of data obtained by scraping webpages of popular news channels like CNN. And the COVID-19 data has been extracted from the largest collection, which is accumulated by John Hopkins University, every day. Then, NLP text processing techniques were used to prepare the dataset and pre-process the text, through which we obtain a labelled dataset. From this, our methodology includes analyzing the obtained dataset.