Twitter Sentiment Analysis of Public Opinions on Work from Home Methodology

The 2020 world pandemic has brought into limelight an already existing and refurbished methodology of working known as ‘Work from Home’. With preventive and curative treatments steadily reaching out most parts of the world now, the existence of Novel Coronavirus and its variants still cannot be denied. When more than 80% of the world was under lockdown and nearly a million people were at a risk of losing their lives, sectors like corporate, education, business and many others, on the recommendation of World Health Organization decided to allow their employees to work from remote locations where they can self-isolate themselves and prevent the spread on this infectious virus thereby some completely while others partially adhering to the work from home methodology. While being at home, people saw social media as a major platform to showcase their views and thinking over the ongoing and post-pandemic scenarios. The main aim of this study is to analyze public sentiments and emotions over Work from Home methodology based upon their views laid out on Twitter by using TextBlob, VADER and various Machine Learning Classification Algorithms to create a conclusive yet generalized picture of the presented positive, negative and neutral aspects of the viewpoints with the objective of identifying the most prominent one. Experimental evaluations show that people, in general, reflect ‘positive to neutral’ opinions towards the Work from Home methodology with Support Vector Classifier classifying the data into their respective sentiment classes with highest possible accuracy of 94.67%.