Deception detection in news reports in the russian language: Lexics and discourse

Abstract

Different language markers can be used to reveal the differences between structures of truthful and deceptive (fake) news. Two experiments are held: The first one is based on lexics level markers, the second one on discourse level is based on rhetorical relations categories (frequencies). Corpus consists of 174 truthful and deceptive news stories in Russian. Support Vector Machines and Random Forest Classifier were used for text classification. The best results for lexical markers we got by using Support Vector Ma-chines with rbf kernel (f-measure 0.65). The model could be developed and be used as a preliminary filter for fake news detection.