Impact of COVID-19 on the Human Personality: An Analysis Based on Document Modeling Using Machine Learning Tools

Abstract

Coronavirus disease of 2019 (COVID-19) has affected the globe terribly. The rapid spread of this virus and the precautionary measures to prevent it have impacted the lives of all human beings around the world in all dimensions. The anxieties over the virus along with the social restrictions have challenged the mental health and might have acute psychological consequences. In this study, our aim is to analyze whether COVID-19 has done any significant changes to very well-known five-factor personality traits of all the humans all over the world from social media text, such as Twitter. We first train and validate five machine learning models on the benchmark essays dataset and then those models are tested on the preprocessed Twitter dataset, consisting of pre_covid and post_covid tweets. The novelty of this study is to analyze and establish the fact that in this short period of time, COVID-19 cannot make very significant changes in the human personality all over the world. We have compared the performances of five machine learning models and what we have found is that the result provided by one model is also justified by the other models.