Fake News Detection Incorporating Emotion Transition in Text

Le résumé fourni par la source

With the rapid development of social media platforms, more and more users are willing to post and comment on news on social platforms, which has also created certain conditions for the massive spread of false news. Therefore, it has become urgent to study methods for false news detection. Some studies have confirmed that using emotional features contained in text can help classifiers detect fake news, but relatively little attention has been paid to the discussion of multiple emotions contained in a message or comment text. Therefore, a method to fuse emotions in text is proposed. The conversion feature detection strategy analyzes the various emotional results contained in the text, uses the conversion characteristics between emotions for modeling analysis, and uses it as an enhanced feature of the text's emotional features to be used by the classifier to detect false news. Experiments were conducted on real data sets, and the results showed that this method is superior to other studies that utilize text emotional features, and can effectively improve the accuracy of detection and achieve good results.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.