PUBLIC SENTIMENT ANALYSIS OF CORONA - VIRSUS SPECIFIC TWEETS USING BERT

Computer Science Engineering

Sentiment analysis is a technique or method that is commonly used to analyze the sentiments or emotions of people. This is a method that is applicable to all occasions and this is also applicable to the global pandemic of the new coronavirus (COVID-19) is driving unprecedented digital conversations on social media. Often expressed through tweets, these conversations provide valuable insight into public feelings, concerns, and attitudes toward the sentiments. In this study, we propose a novel approach to explore public sentiment towards COVID-19 by using the bidirectional encoder representation from the transformer (BERT) model. Our research involves collecting an extensive dataset of coronavirus-specific tweets from social media platforms covering a key period. We then per-process and clean the data to remove unnecessary noise and information. Then, we use this prepared dataset to fine-tune the BERT model, enabling us to understand the nuances and context of discussions about COVID. We use this refined model to analyze sentiment on a collection of tweets. Sentiment analysis provides a comprehensive view of public emotional responses to different aspects of the epidemic by classifying tweets into positive, negative, and neutral sentiments Furthermore, we use natural language processing techniques to advanced use to extract key topics and trends from tweets, so that public discourse A deeper understanding is possible. Furthermore, we examine temporal changes in sentiment and information during the pandemic, identifying significant changes in public sentiment as events unfold. These longitudinal surveys help monitor changing public perceptions and concerns, which can be valuable for policymakers and health professionals.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.