Sentimental Analysis On Covid-19 Tweets using Bidirectional Encoder Representations Transformers

Abstract

The Coronavirus epidemic has wreaked havoc on countries all over the world. People from all over the Globe have flocked to social network to voice their thoughts and feelings about the situation that has reached epidemic proportions. The need of this paper is to exemplify how social media users feel about COVID-19 in an extremely brief span of time, Twitter saw an unusual surge in tweets on the new Coronavirus. This study presents a global Sentiment Analysis of tweets related to COVID-19, as well as how people’s sentiment in multiple nations has varied. The research study focuses on a period of time from march 2020 to april 2020. In the Sentiment Analysis, we fed dataset to different algorithms and estimate the best performance among them. As in secondly we also found the reliability on BERT model. Comparatively, BERT gave foremost accuracy amidst all. Besides, the accuracy of mentioned algorithms are well represented.