Topic detection and sentiment analysis in Twitter content related to COVID-19 from Brazil and the USA - Dataset

About

This is a dataset containing twitter message id's, which we use to analyze messages related to COVID-19. The messages were collected between April 17,2020 through August 08, 2020, and each dataset has approximately 3M id's.

Usage

Download and unzip the files. The language can be identified by the substring pt or en in the file name.

You can use these IDs to download the full tweet objects. You can do this using a script like this.

If you use this data, please quote our work:

@article{GARCIA2021107057,
  title = {Topic detection and sentiment analysis in Twitter content related to COVID-19 from Brazil and the USA},
  journal = {Applied Soft Computing},
  volume = {101},
  pages = {107057},
  year = {2021},
  issn = {1568-4946},
  doi = {https://doi.org/10.1016/j.asoc.2020.107057},
  url = {https://www.sciencedirect.com/science/article/pii/S1568494620309959},
  author = {Klaifer Garcia and Lilian Berton},
  keywords = {COVID-19, Twitter, Topic detection, Sentiment analysis, Portuguese language, English language},
}