Deep Reinforcement Learning for Dialogue Generation

TL;DR

This paper introduces a deep reinforcement learning approach for dialogue generation that optimizes for long-term conversational quality, resulting in more coherent, informative, and engaging chatbot responses.

Contribution

It presents a novel integration of deep reinforcement learning with dialogue modeling, focusing on long-term reward optimization for improved conversational coherence.

Findings

Generated responses are more diverse and less repetitive.

The model sustains longer and more engaging conversations.

Human evaluations favor the proposed method over baseline models.

Abstract

Recent neural models of dialogue generation offer great promise for generating responses for conversational agents, but tend to be shortsighted, predicting utterances one at a time while ignoring their influence on future outcomes. Modeling the future direction of a dialogue is crucial to generating coherent, interesting dialogues, a need which led traditional NLP models of dialogue to draw on reinforcement learning. In this paper, we show how to integrate these goals, applying deep reinforcement learning to model future reward in chatbot dialogue. The model simulates dialogues between two virtual agents, using policy gradient methods to reward sequences that display three useful conversational properties: informativity (non-repetitive turns), coherence, and ease of answering (related to forward-looking function). We evaluate our model on diversity, length as well as with human judges, showing that the proposed algorithm generates more interactive responses and manages to foster a more sustained conversation in dialogue simulation. This work marks a first step towards learning a neural conversational model based on the long-term success of dialogues.