A Cloud-based Framework for COVID-19 Media Classification, Information Extraction, and Trends Analysis

Abstract

The coronavirus COVID-19 pandemic has become the center of concern worldwide and hence the focus of media attention. Checking the coronavirus-related news and updates has become a daily routine of everyone. Hence, news processing and analytics become key solutions to harvest the real value of this massive amount of news. This conscious growth of published news about COVID-19 makes it hard for a variety of audiences to navigate through, analyze, and select the most important news (e.g., relevant information about the pandemic, its evolution, the vital precautions, and the necessary interventions). This can be realized using current and emerging technologies including Cloud computing, Artificial Intelligence (AI) and Deep Learning (DL). In this paper, we propose a framework to analyze the massive amount of public Covid-19 media reports over the Cloud. This framework encompasses four modules, including text preprocessing, deep learning, and machine learning-based news information extraction, and recommendation. We conducted experiments to evaluate three modules of our framework and the results we have obtained prove that combining derived information from the news reports provides the policymakers, health authorities, and the public, a complete picture of the way this virus is proliferating. Analyzing this data swiftly is a powerful tool to provide imperative answers to questions that are relevant to public health.