Adversarial Domain Adaptation for Variational Neural Language Generation in Dialogue Systems
This addresses the problem of limited labeled data in target domains for dialogue system developers, though it appears incremental as it builds on existing adversarial and variational methods.
The paper tackles domain adaptation for natural language generation in dialogue systems by proposing an adversarial training procedure for a variational encoder-decoder model, showing it effectively leverages source domain knowledge to adapt to a target domain with limited data.
Domain Adaptation arises when we aim at learning from source domain a model that can per- form acceptably well on a different target domain. It is especially crucial for Natural Language Generation (NLG) in Spoken Dialogue Systems when there are sufficient annotated data in the source domain, but there is a limited labeled data in the target domain. How to effectively utilize as much of existing abilities from source domains is a crucial issue in domain adaptation. In this paper, we propose an adversarial training procedure to train a Variational encoder-decoder based language generator via multiple adaptation steps. In this procedure, a model is first trained on a source domain data and then fine-tuned on a small set of target domain utterances under the guidance of two proposed critics. Experimental results show that the proposed method can effec- tively leverage the existing knowledge in the source domain to adapt to another related domain by using only a small amount of in-domain data.