Replication Files for the Article "When Tweets Get Viral - A Deep Learning Approach for Stance Analysis of Covid-19 Vaccines Tweets by Brazilian Political Elites", by Lorena Barberia, Pedro Schmal and Norton Roman

We made available the following replication files:

BERT+KFOLD(2020).ipynb: The code and results of the model trained only on 2020 tweets.

BERT+KFOLD(2021).ipynb: The code and results of the model trained only on 2021 tweets.

BERT+KFOLD(2020-21).ipynb: The code and results of the model trained on the tweets of both years.

Codebook: Additionally, we provide the codebook with informations on the dataset building process, classification rules and some additional content.

Slides: Presentation Slides, STIL 2023.

Note: The models were trained using Google Colaboratory GPU.

The corpus used on the training can be found on the following site

BARBERIA, Lorena Guadalupe; SCHMALZ, Pedro Henrique de Santana; ROMAN, Norton Trevisan. When Tweets Get Viral - A Deep Learning Approach for Stance Analysis of Covid-19 Vaccines Tweets by Brazilian Political Elites. In: SIMPÓSIO BRASILEIRO DE TECNOLOGIA DA INFORMAÇÃO E DA LINGUAGEM HUMANA (STIL), 14. , 2023, Belo Horizonte/MG. Anais [...]. Porto Alegre: Sociedade Brasileira de Computação, 2023 . p. 104-114. DOI: https://doi.org/10.5753/stil.2023.233961.

Shield:

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.