Task-Optimized Adapters for an End-to-End Task-Oriented Dialogue System

TL;DR

This paper introduces task-optimized adapters for end-to-end task-oriented dialogue systems, enabling lightweight, independent task learning, and improves dialogue state tracking and response generation with reinforcement learning, achieving state-of-the-art results.

Contribution

The paper proposes a model-agnostic adapter-based approach for task-specific modules in dialogue systems, reducing complexity and enhancing performance without prompt-tuning.

Findings

Achieves state-of-the-art dialogue state tracking accuracy on MultiWOZ.

Demonstrates competitive overall performance on the MultiWOZ benchmark.

Enhances response generation quality through reinforcement learning.

Abstract

Task-Oriented Dialogue (TOD) systems are designed to carry out specific tasks by tracking dialogue states and generating appropriate responses to help users achieve defined goals. Recently, end-to-end dialogue models pre-trained based on large datasets have shown promising performance in the conversational system. However, they share the same parameters to train tasks of the dialogue system (NLU, DST, NLG), so debugging each task is challenging. Also, they require a lot of effort to fine-tune large parameters to create a task-oriented chatbot, making it difficult for non-experts to handle. Therefore, we intend to train relatively lightweight and fast models compared to PLM. In this paper, we propose an End-to-end TOD system with Task-Optimized Adapters which learn independently per task, adding only small number of parameters after fixed layers of pre-trained network. We also enhance the performance of the DST and NLG modules through reinforcement learning, overcoming the learning curve that has lacked at the adapter learning and enabling the natural and consistent response generation that is appropriate for the goal. Our method is a model-agnostic approach and does not require prompt-tuning as only input data without a prompt. As results of the experiment, our method shows competitive performance on the MultiWOZ benchmark compared to the existing end-to-end models. In particular, we attain state-of-the-art performance on the DST task of 2.2 dataset.