Deep Reinforcement Learning for Inquiry Dialog Policies with Logical Formula Embeddings

TL;DR

This paper introduces a novel deep reinforcement learning approach combined with logical formula embeddings to learn inquiry dialog policies, demonstrating effectiveness comparable or superior to rule-based methods.

Contribution

It is the first to apply deep reinforcement learning with logical formula embeddings to inquiry dialog policy learning, enhancing effectiveness.

Findings

DRL with logical embeddings outperforms rule-based methods

Logical formula embeddings improve dialog policy learning

The combined approach is as effective or better than existing methods

Abstract

This paper is the first attempt to learn the policy of an inquiry dialog system (IDS) by using deep reinforcement learning (DRL). Most IDS frameworks represent dialog states and dialog acts with logical formulae. In order to make learning inquiry dialog policies more effective, we introduce a logical formula embedding framework based on a recursive neural network. The results of experiments to evaluate the effect of 1) the DRL and 2) the logical formula embedding framework show that the combination of the two are as effective or even better than existing rule-based methods for inquiry dialog policies.