Detection of conspiracy propagators using psycho-linguistic characteristics

The rise of social media has offered a fast and easy way for the propagation conspiracy theories other types disinformation. Despite research attention that received, fake news detection remains an open problem users keep sharing articles contain false statements but which they consider real. In this article, we focus on role in is specific type First, compare profile psycho-linguistic patterns online tend to propagate posts support those who refute them. To end, perform comparative analysis over various profile, psychological linguistic characteristics using texts share about theories. Then, effectiveness predicting whether user propagator or not. addition, propose ConspiDetector, model based convolutional neural network (CNN) combines word embeddings with extracted from tweets detect propagators. results show ConspiDetector can improve performance detecting propagators by 8.82% compared CNN baseline regard F1-metric.