On the Robustness of Transformer‐Based Models to Different Linguistic Perturbations: A Case of Study in Irony Detection | Publicación