COVID-19 fake news detection using attention-based transformer

Konferenz: ICMLCA 2021 - 2nd International Conference on Machine Learning and Computer Application
17.12.2021 - 19.12.2021 in Shenyang, China

Tagungsband: ICMLCA 2021

Seiten: 7Sprache: EnglischTyp: PDF

Persönliche VDE-Mitglieder erhalten auf diesen Artikel 10% Rabatt

Autoren:
Xi, Jiayang (School of Information Science and Technology, Nanjing Forestry University, Nanjing, Jiangsu, China)
Zhang, Chaokai (School of Engineering, University of Connecticut, Storrs, Connecticut, USA)

Inhalt:
The worldwide epidemic, COVID-19, has resulted in the deaths of millions of people, and now coronavirus has gone globally. Meanwhile, misinformation about COVID-19 spread throughout the social networks like a virus, affecting social order during the pandemic. In this paper, three techniques, including attention-based transformer, CNN, and LSTM models, construct a COVID-19 fake news detector. The model performance parameters of accuracy, F1 score, AUC score, and training time involve evaluating the fake news detector. Furthermore, confusion matrix and ROC curve analysis have been employed to further analyze the attention-based transformer model's performance. Following the findings, the attention-based transformer model has shown its advantages through its high accuracy of 75.9% and AUC score of 0.774.