Generating Informative And Diverse Conversational Responses Via Adversarial Information Maximization

Authors:
Yizhe Zhang Microsoft Research
Michel Galley Microsoft Research
Jianfeng Gao Microsoft Research, Redmond, WA
Zhe Gan Microsoft
Xiujun Li Microsoft Research Redmond
Chris Brockett Microsoft Research AI
Bill Dolan Microsoft

Introduction:

Responses generated by neural conversational models tend to lack informativeness and diversity.The authors present Adversarial Information Maximization (AIM), an adversarial learning framework that addresses these two related but distinct problems.

Abstract:

Responses generated by neural conversational models tend to lack informativeness and diversity. We present Adversarial Information Maximization (AIM), an adversarial learning framework that addresses these two related but distinct problems. To foster response diversity, we leverage adversarial training that allows distributional matching of synthetic and real responses. To improve informativeness, our framework explicitly optimizes a variational lower bound on pairwise mutual information between query and response. Empirical results from automatic and human evaluations demonstrate that our methods significantly boost informativeness and diversity.

You may want to know: