Does fake news have feelings?

Laroca Mendes Pinto, Herbert; Rocio, Vitor; Cunha, António

http://hdl.handle.net/10400.2/19425

Use this identifier to reference this record.

Name:	Description:	Size:	Format:
1-s2.0-S1877050924016363-main.pdf		793.26 KB	Adobe PDF	Download

Send Feedback

Authors

Laroca Mendes Pinto, Herbert

Rocio, Vitor

Cunha, António

Abstract(s)

Fake news spreads rapidly, creating issues and making detection harder. The purpose of this study is to determine if fake news contains sentiment polarity (positive or negative), identify the polarity of sentiment present in their textual content and determine whether sentiment polarity is a reliable indication of fake news. For this, we use a deep learning model called BERT (Bidirectional Encoder Representations from Transformers), trained on a sentiment polarity dataset to classify the polarity of sentiments from a dataset of true and fake news. The findings show that sentiment polarity is not a reliable single feature for recognizing false news correctly and must be combined with other parameters to improve classification accuracy.