Ensemble-Based Deep Reinforcement Learning for Chatbots

TL;DR

This paper introduces an ensemble-based deep reinforcement learning approach for training chatbots that can generate human-like conversations without manual data labeling, demonstrating improved performance and human-like dialogue quality.

Contribution

The paper presents a novel ensemble-based DRL method for chatbots that learns from raw text data, enabling specialized agents and improving dialogue quality over single-agent models.

Findings

Ensemble of agents outperforms single-agent models.

Near human-like dialogue policies can be learned.

Dialogue rewards correlate with human judgments.

Abstract

Trainable chatbots that exhibit fluent and human-like conversations remain a big challenge in artificial intelligence. Deep Reinforcement Learning (DRL) is promising for addressing this challenge, but its successful application remains an open question. This article describes a novel ensemble-based approach applied to value-based DRL chatbots, which use finite action sets as a form of meaning representation. In our approach, while dialogue actions are derived from sentence clustering, the training datasets in our ensemble are derived from dialogue clustering. The latter aim to induce specialised agents that learn to interact in a particular style. In order to facilitate neural chatbot training using our proposed approach, we assume dialogue data in raw text only -- without any manually-labelled data. Experimental results using chitchat data reveal that (1) near human-like dialogue policies can be induced, (2) generalisation to unseen data is a difficult problem, and (3) training an ensemble of chatbot agents is essential for improved performance over using a single agent. In addition to evaluations using held-out data, our results are further supported by a human evaluation that rated dialogues in terms of fluency, engagingness and consistency -- which revealed that our proposed dialogue rewards strongly correlate with human judgements.