Dialogue Act Classification in Domain-Independent Conversations Using a Deep Recurrent Neural Network.

This paper applies a deep long-short term memory (LSTM) structure to classify dialogue acts in open-domain conversations.