A Copy-Augmented Sequence-to-Sequence Architecture Gives Good Performance on Task-Oriented Dialogue
This provides a simpler, effective approach for conversational agents in domain-specific tasks, though it is incremental as it builds on existing sequence-to-sequence and copy mechanism methods.
The paper tackles task-oriented dialogue by bypassing explicit state representations, using a sequence-to-sequence model with a copy mechanism, which outperforms complex memory-augmented models by 7% in per-response generation and matches state-of-the-art on DSTC2.
Task-oriented dialogue focuses on conversational agents that participate in user-initiated dialogues on domain-specific topics. In contrast to chatbots, which simply seek to sustain open-ended meaningful discourse, existing task-oriented agents usually explicitly model user intent and belief states. This paper examines bypassing such an explicit representation by depending on a latent neural embedding of state and learning selective attention to dialogue history together with copying to incorporate relevant prior context. We complement recent work by showing the effectiveness of simple sequence-to-sequence neural architectures with a copy mechanism. Our model outperforms more complex memory-augmented models by 7% in per-response generation and is on par with the current state-of-the-art on DSTC2.