Towards End-to-End Reinforcement Learning of Dialogue Agents for Information Access

TL;DR

This paper introduces KB-InfoBot, a neural dialogue system that uses a differentiable soft retrieval method for knowledge base access, enabling end-to-end training and improved task success in goal-oriented dialogues.

Contribution

It replaces symbolic KB queries with a soft posterior, allowing fully neural, end-to-end training of dialogue agents for knowledge access.

Findings

Higher task success rate with soft retrieval

Effective training from user feedback

Applicable to personalized dialogue agents

Abstract

This paper proposes KB-InfoBot -- a multi-turn dialogue agent which helps users search Knowledge Bases (KBs) without composing complicated queries. Such goal-oriented dialogue agents typically need to interact with an external database to access real-world knowledge. Previous systems achieved this by issuing a symbolic query to the KB to retrieve entries based on their attributes. However, such symbolic operations break the differentiability of the system and prevent end-to-end training of neural dialogue agents. In this paper, we address this limitation by replacing symbolic queries with an induced "soft" posterior distribution over the KB that indicates which entities the user is interested in. Integrating the soft retrieval process with a reinforcement learner leads to higher task success rate and reward in both simulations and against real users. We also present a fully neural end-to-end agent, trained entirely from user feedback, and discuss its application towards personalized dialogue agents. The source code is available at https://github.com/MiuLab/KB-InfoBot.