Abstract:End-to-end multi-task dialogue systems are usually designed with separate modules for the dialogue pipeline. Among these, the policy module is essential for deciding what to do in response to user input. This policy is trained by reinforcement learning algorithms by taking advantage of an environment in which an agent receives feedback in the form of a reward signal. The current dialogue systems, however, only provide meagre and simplistic rewards. Investigating intrinsic motivation reinforcement learning algorithms is the goal of this study. Through this, the agent can quickly accelerate training and improve its capacity to judge the quality of its actions by teaching it an internal incentive system. In particular, we adapt techniques for random network distillation and curiosity-driven reinforcement learning to measure the frequency of state visits and encourage exploration by using semantic similarity between utterances. Experimental results on MultiWOZ, a heterogeneous dataset, show that intrinsic motivation-based debate systems outperform policies that depend on extrinsic incentives. By adopting random network distillation, for example, which is trained using semantic similarity between user-system dialogues, an astounding average success rate of 73% is achieved. This is a significant improvement over the baseline Proximal Policy Optimization (PPO), which has an average success rate of 60%. In addition, performance indicators such as booking rates and completion rates show a 10% rise over the baseline. Furthermore, these intrinsic incentive models help improve the system's policy's resilience in an increasing amount of domains. This implies that they could be useful in scaling up to settings that cover a wider range of domains.

Learning Dialogue Policy Efficiently Through Dyna Proximal Policy Optimization.

Gaussian Process Based Deep Dyna-Q Approach for Dialogue Policy Learning.

Discriminative Deep Dyna-Q: Robust Planning for Dialogue Policy Learning

Dynamic Reward-Based Dueling Deep Dyna-Q: Robust Policy Learning in Noisy Environments

Emotion-sensitive deep dyna-Q learning for task-completion dialogue policy learning

Budgeted Policy Learning for Task-Oriented Dialogue Systems

Deep Dyna-Q: Integrating Planning for Task-Completion Dialogue Policy Learning.

Agent-Aware Dropout DQN for Safe and Efficient On-line Dialogue Policy Learning.

Switch-Based Active Deep Dyna-Q: Efficient Adaptive Planning for Task-Completion Dialogue Policy Learning.

Efficient Dialogue Complementary Policy Learning Via Deep Q-network Policy and Episodic Memory Policy.

Automatic Curriculum Learning with Over-repetition Penalty for Dialogue Policy Learning

Proximal Policy Optimization with Mixed Distributed Training

Interacting with Non-Cooperative User: A New Paradigm for Proactive Dialogue Policy

Demonstration-Based Proximal Policy Optimization with Action Guidance

Simulation-Free Hierarchical Latent Policy Planning for Proactive Dialogues

An Improved Proximal Policy Optimization Algorithm for Autonomous Driving Decision-Making

Enhancing End-to-End Multi-Task Dialogue Systems: A Study on Intrinsic Motivation Reinforcement Learning Algorithms for Improved Training and Adaptability

Goal-Oriented Dialogue Policy Learning from Failures

Decomposed Deep Q-Network for Coherent Task-Oriented Dialogue Policy Learning

Policy Networks with Two-Stage Training for Dialogue Systems

Guided Dialog Policy Learning: Reward Estimation for Multi-Domain Task-Oriented Dialog