Few-shot Natural Language Generation for Task-Oriented Dialog
This work addresses the challenge of data scarcity for natural language generation in task-oriented dialog systems, enabling better generalization to new domains with minimal labeled data.
The authors tackled the problem of generating natural language responses in task-oriented dialog systems with limited labeled data by introducing FewShotWoz, a benchmark for few-shot learning, and SC-GPT, a model that pre-trains on large annotated corpora and fine-tunes with few domain-specific labels. Experiments showed that SC-GPT significantly outperforms existing methods on FewShotWoz and Multi-Domain-WOZ datasets, as measured by automatic metrics and human evaluations.
As a crucial component in task-oriented dialog systems, the Natural Language Generation (NLG) module converts a dialog act represented in a semantic form into a response in natural language. The success of traditional template-based or statistical models typically relies on heavily annotated data, which is infeasible for new domains. Therefore, it is pivotal for an NLG system to generalize well with limited labelled data in real applications. To this end, we present FewShotWoz, the first NLG benchmark to simulate the few-shot learning setting in task-oriented dialog systems. Further, we develop the SC-GPT model. It is pre-trained on a large set of annotated NLG corpus to acquire the controllable generation ability, and fine-tuned with only a few domain-specific labels to adapt to new domains. Experiments on FewShotWoz and the large Multi-Domain-WOZ datasets show that the proposed SC-GPT significantly outperforms existing methods, measured by various automatic metrics and human evaluations.