IDENTIFYING EMOTIONS DURING COVID-19 USING TOPIC MODELLING APPROACH

Authors

  • Rajkumar Pillay, et. al.

Abstract

The purpose of this paper was to study the emotions of large population under the unique situation like the COVID-19,

a). How to capture the emotions of large population without the direct observation of the respondent and his facial expression while the distressed situation is ongoing?

b). Will the basic emotions remain the same or will they be different while the respondents are not directly involved?

Resent researchers propose that “emotional experiences also can be investigated using qualitative

methods such as the coding of written or oral narratives” such as sentiment analysis of Tweets. Therefore, this study was carried out by identifying #COVID-19 Tweets during first seven-day stretch of August-2020 utilizing supposition study and point of demonstrating utilizing Latent Dirichlet Allocation(LDA) post preparing. The study reasoned that the data stream was exact and dependable identified with COVID-19 outbreak with least deception.

From the data analysis we found that the oral narratives of large population related to #COVID-19 have represented the basic emotions same as individual facial expression or self-reported questionnaires that are; “fear”, “sadness”, “joy”, “surprise”, “anger”, and “disgust”,  Therefore, we can conclude that the Topic Modelling Approach” or  LDA methodology can be used for twitter hashtags analysis with equal reliability as that of observation and self-reporting studies. In addition, the advantage can be a quicker process with a large number of respondents that gives better reliability of the data. We also found that using Topic Model Approach, one aspect of measurement may serve as a proxy for others

Downloads

Published

2021-03-28

How to Cite

et. al., R. P. . (2021). IDENTIFYING EMOTIONS DURING COVID-19 USING TOPIC MODELLING APPROACH. International Journal of Modern Agriculture, 10(2), 571 - 580. Retrieved from http://www.modern-journals.com/index.php/ijma/article/view/775

Issue

Section

Articles