Application of Supervised Machine Learning techniques for COVID- 19 Text document Categorization

Abstract

The COVID-19 pandemic forced the
research community to discover different methods,
ideas, and medicines to handle the pandemic. The
research articles in COVID-19 literature have been
growing exponentially, and manual classification of
these articles is an impossible task. Therefore, automatic
extraction and classification of COVID-19 related
articles from the vast COVID-19 literature emerge as a
significant task. Hence, thisthesis implements the vital
Machine Learning (ML) algorithms like decision tree, knearest
neighbourhood, Rocchio, ridge, passiveaggressive,
multinomial naïveBayes, Bernoulli naïve
Bayes, support vector machine, and artificial neural
network classifiers such as perceptron, random gradient
descent, and Backpropagation neural network in
automatic classification of COVID-19 text documents
on benchmark PubMed Abstract dataset. Finally, the
performance of all the said constitutional classifiers are
compared and evaluated utilizing the well-defined
metrics like accuracy, error rate, precision, recall, and f-measure.