Generating Relevant and Coherent Dialogue Responses using Self-separated Conditional Variational AutoEncoders

TL;DR

This paper introduces SepaCVAE, a novel model that improves relevance and coherence in dialogue response generation by regularizing latent variables through group information, enhancing diversity and informativeness.

Contribution

SepaCVAE incorporates group-based regularization to better align latent variables with context semantics, addressing incoherence issues in CVAE for dialogue generation.

Findings

Significantly improves response relevance and coherence.

Maintains high diversity and informativeness of responses.

Outperforms baseline models on open-domain dialogue datasets.

Abstract

Conditional Variational AutoEncoder (CVAE) effectively increases the diversity and informativeness of responses in open-ended dialogue generation tasks through enriching the context vector with sampled latent variables. However, due to the inherent one-to-many and many-to-one phenomena in human dialogues, the sampled latent variables may not correctly reflect the contexts' semantics, leading to irrelevant and incoherent generated responses. To resolve this problem, we propose Self-separated Conditional Variational AutoEncoder (abbreviated as SepaCVAE) that introduces group information to regularize the latent variables, which enhances CVAE by improving the responses' relevance and coherence while maintaining their diversity and informativeness. SepaCVAE actively divides the input data into groups, and then widens the absolute difference between data pairs from distinct groups, while narrowing the relative distance between data pairs in the same group. Empirical results from automatic evaluation and detailed analysis demonstrate that SepaCVAE can significantly boost responses in well-established open-domain dialogue datasets.