How to Evaluate the Next System: Automatic Dialogue Evaluation from the Perspective of Continual Learning

TL;DR

This paper proposes a continual learning approach to automatically evaluate dialogue systems, enabling a neural evaluator to adapt to new systems efficiently without retraining from scratch.

Contribution

It introduces a continual learning method for dialogue evaluation, reducing resource costs and maintaining performance across multiple dialogue systems.

Findings

Continual evaluator performs comparably to retrained evaluators.

Requires significantly fewer resources than rebuilding evaluators.

Effective in adapting to new dialogue systems over time.

Abstract

Automatic dialogue evaluation plays a crucial role in open-domain dialogue research. Previous works train neural networks with limited annotation for conducting automatic dialogue evaluation, which would naturally affect the evaluation fairness as dialogue systems close to the scope of training corpus would have more preference than the other ones. In this paper, we study alleviating this problem from the perspective of continual learning: given an existing neural dialogue evaluator and the next system to be evaluated, we fine-tune the learned neural evaluator by selectively forgetting/updating its parameters, to jointly fit dialogue systems have been and will be evaluated. Our motivation is to seek for a lifelong and low-cost automatic evaluation for dialogue systems, rather than to reconstruct the evaluator over and over again. Experimental results show that our continual evaluator achieves comparable performance with reconstructing new evaluators, while requires significantly lower resources.