Less is More: Mitigate Spurious Correlations for Open-Domain Dialogue Response Generation Models by Causal Discovery
This addresses the issue of unreliable dialogue systems for users by mitigating spurious correlations, though it is incremental as it builds on existing causal discovery techniques.
The paper tackles the problem of spurious correlations in open-domain dialogue response generation models, which cause irrelevant and generic responses, by proposing a model-agnostic method using causal discovery and a conditional independence classifier, resulting in significant improvements in relevance, informativeness, and fluency over baselines.
In this paper, we conduct the first study on spurious correlations for open-domain response generation models based on a corpus CGDIALOG curated in our work. The cur rent models indeed suffer from spurious correlations and have a tendency of generating irrelevant and generic responses. Inspired by causal discovery algorithms, we propose a novel model-agnostic method for training and inference of response generation model using a conditional independence classifier. The classifier is trained by a constrained self-training method, coined CONSTRAIN, to overcome data scarcity. The experimental results based on both human and automatic evaluation show that our method significantly outperforms the competitive baselines in terms of relevance, informativeness, and fluency.