In this project, we try to detect metaphors using recurrent neural networks. Particularly, we analyze how metaphors are used in the context of fake news