Covid-19 Online Articles Dataset

Published: 24 March 2021| Version 2 | DOI: 10.17632/r6nn5s37tp.2


This dataset is a collection of articles about Covid-19 published online from May 2020 to September 2020 and stored as a CSV file. The primary providers of these articles are,, and The dataset contains two columns (text and sentiment). The text column contains text from the articles to which a label applies. The sentiment column contains either the value 1 (positive class) for text with positive sentiment or the value 0 (negative class) for text with negative sentiment. The model used will be published in one of the journals later and will be found on my profile with the title: 'Sentiment Analysis of Covid-19 Articles; The Impact of Bidirectional Layer on Long Short-Term Memory (LSTM).'



University of Electronic Science and Technology of China


Natural Language Processing, Sentiment Analysis