Building End-To-End Dialogue Systems Using Generative Hierarchical Neural Network Models
This addresses the problem of creating more realistic and flexible dialogue systems for AI applications, but it is incremental as it extends existing hierarchical models to the dialogue domain.
The authors tackled building open-domain conversational dialogue systems using generative hierarchical neural networks, achieving competitive performance with state-of-the-art neural language models and back-off n-gram models, and improved it by bootstrapping from larger question-answer corpora and pretrained word embeddings.
We investigate the task of building open domain, conversational dialogue systems based on large dialogue corpora using generative models. Generative models produce system responses that are autonomously generated word-by-word, opening up the possibility for realistic, flexible interactions. In support of this goal, we extend the recently proposed hierarchical recurrent encoder-decoder neural network to the dialogue domain, and demonstrate that this model is competitive with state-of-the-art neural language models and back-off n-gram models. We investigate the limitations of this and similar approaches, and show how its performance can be improved by bootstrapping the learning from a larger question-answer pair corpus and from pretrained word embeddings.