A Fine-Grained Study of Interpretability of Convolutional Neural Networks for Text Classification | Publicación