Knowledge-Grounded Response Generation with Deep Attentional Latent-Variable Model
This work addresses a key limitation in conversational AI for real-world applications, though it appears incremental as it builds on existing variational models.
The paper tackled the problem of generating uninformative responses in end-to-end dialogue systems by introducing a variational generation model with joint attention on dialogue contexts and external knowledge, resulting in more diverse and informative utterances.
End-to-end dialogue generation has achieved promising results without using handcrafted features and attributes specific for each task and corpus. However, one of the fatal drawbacks in such approaches is that they are unable to generate informative utterances, so it limits their usage from some real-world conversational applications. This paper attempts at generating diverse and informative responses with a variational generation model, which contains a joint attention mechanism conditioning on the information from both dialogue contexts and extra knowledge.