An Integrated Feature Engineering Approach to Identify Falsity in COVID-19 Information

Abstract

There is huge flow of data in social media which obviously causes a rise in the amount of information. This is more evident particularly in the information regarding COVID-19 pandemic. Despite of the information being periodically updated on regular basis, lot of disturbances still exists. Countering misinformation is a mammoth task. The misinformation during this time should be monitored properly as it creates havoc in the society if it's left unattended. So identifying fake information is the main objective of this investigation. The analysis is based on a dataset of 1100 posts related to COVID 19 collected from social media. We propose an Integrated Feature Engineering (IFE) approach which uses a combination of content based and count vectorization methods. This study focuses on applying the proposed approach and gives a comparison with in individual methods and finally assesses the performance of different machine learning classifiers for classifying fake news.