A Conditional Variational Framework for Dialog Generation

TL;DR

This paper introduces a conditional variational framework for dialog generation that allows control over responses based on specific attributes like sentiment or genericness, improving response relevance and personalization.

Contribution

It presents a novel framework that models dialog states separately and enables attribute-controlled response generation, enhancing controllability and personalization in open-domain dialog systems.

Findings

Generated responses align with specified attributes

Model effectively captures personal features in dialog states

Framework tested on sentiment and genericness scenarios

Abstract

Deep latent variable models have been shown to facilitate the response generation for open-domain dialog systems. However, these latent variables are highly randomized, leading to uncontrollable generated responses. In this paper, we propose a framework allowing conditional response generation based on specific attributes. These attributes can be either manually assigned or automatically detected. Moreover, the dialog states for both speakers are modeled separately in order to reflect personal features. We validate this framework on two different scenarios, where the attribute refers to genericness and sentiment states respectively. The experiment result testified the potential of our model, where meaningful responses can be generated in accordance with the specified attributes.