Classification of COVID-19 Tweets using Bidirectional Encoder Representation for Transformers(BERT)
Year: 2021
Type: dissertation
Abstract: The outbreak of COVID-19 in the later part of 2019 caused a lot of panic and led to the\nloss of millions of lives. Much of the chaos could have been avoided if the spread was\ndetected in the early stages of the outbreak. Also had there been adequate information\nabout its mode of transmission, prevention measures and symptoms, the outbreak could\nhave been controlled. In this thesis we perform a classification of COVID-19 tweets using\nBERT.\nBERT is a deep learning algorithm that is designed using transformers. It is broadly\nused on text data in natural language processing. We modify the BERT architecture for\nCOVID-19 tweet classification. We also show how to train the algorithm to identify our\ntweets using biomedical literature abstracts as an alternate data source.\nWe discovered that the BERT model performs unsatisfactorily in our tests in comparison\nto our baseline model, logistic regression. We also learned that the BERT model requires\na large amount of data for training. This is despite the fact that it has been pre-trained.\nWe also discovered that training the model with a combination of tweets and literature\nabstracts improves its performance as opposed to training it with only literature abstracts.
Author: Opare, Samuel
Language: English
Topic: Topic Modeling
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7036900309