Post-Pandemic Sentiment Analysis Based on Twitter Data Using Deep Learning

Since January 2020, COVID-19 has been spreading over the world and has been declared a pandemic. Nation and society are growing scared of it as fresh instances and mass deaths increase daily. India was one of the major countries to suffer the consequences of COVID-19 during that phase as multiple waves hit India. Many social media channels were being used by people from all over the country to discuss this pandemic and its aftereffects. One of the most popular ways to share opinions or judgments today is through social media. Therefore, machines are continuously being developed to analyze what people post on social networking sites like Twitter, Facebook, Instagram, and other platforms thanks to advancements in current computing technology. Based on their mood, these ideas or points of view can be grouped and examined. In this paper, we used tweets collected from Twitter to analyze the sentiment that people conveyed on social media after the second wave of Corona Virus. The sentiment of the tweets has been divided into five categories: "Strongly Negative", "Negative", "Neutral", "Positive" and "Strongly Positive". First, we classify data using Python’s Vader. We have trained a model using our own labeled dataset and evaluated its performance using Long Short-Term Memory (LSTM), and Convolutional Neural Network (CNN).