Interpreting Neural Networks to Improve Politeness Comprehension

Malika Aubakirova, Mohit Bansal

We present an interpretable neural network approach to predicting and understanding politeness in natural language requests. Our models are based on simple convolutional neural net…