Progressive Open-Domain Response Generation with Multiple Controllable Attributes
This addresses the need for enhanced diversity in dialogue systems for users, though it appears incremental by building on existing controllable attribute methods.
The paper tackles the problem of generating diverse responses in open-domain dialogue systems by enabling control over multiple attributes, proposing a Progressively trained Hierarchical Encoder-Decoder (PHED) that significantly outperforms state-of-the-art models and produces more diverse responses as expected.
It is desirable to include more controllable attributes to enhance the diversity of generated responses in open-domain dialogue systems. However, existing methods can generate responses with only one controllable attribute or lack a flexible way to generate them with multiple controllable attributes. In this paper, we propose a Progressively trained Hierarchical Encoder-Decoder (PHED) to tackle this task. More specifically, PHED deploys Conditional Variational AutoEncoder (CVAE) on Transformer to include one aspect of attributes at one stage. A vital characteristic of the CVAE is to separate the latent variables at each stage into two types: a global variable capturing the common semantic features and a specific variable absorbing the attribute information at that stage. PHED then couples the CVAE latent variables with the Transformer encoder and is trained by minimizing a newly derived ELBO and controlled losses to produce the next stage's input and produce responses as required. Finally, we conduct extensive evaluations to show that PHED significantly outperforms the state-of-the-art neural generation models and produces more diverse responses as expected.