Domain Adaptive Dialog Generation via Meta Learning

Abstract

Domain adaptation is an essential task in dialog system building because there are so many new dialog tasks created for different needs every day. Collecting and annotating training data for these new tasks is costly since it involves real user interactions. We propose a domain adaptive dialog generation method based on meta-learning (DAML). DAML is an end-to-end trainable dialog system model that learns from multiple rich-resource tasks and then adapts to new domains with minimal training samples. We train a dialog system model using multiple rich-resource single-domain dialog data by applying the model-agnostic meta-learning algorithm to dialog domain. The model is capable of learning a competitive dialog system on a new domain with only a few training examples in an efficient manner. The two-step gradient updates in DAML enable the model to learn general features across multiple tasks. We evaluate our method on a simulated dialog dataset and achieve state-of-the-art performance, which is generalizable to new tasks.

arXiv commentAccepted as a long paper in ACL 2019

References in corpus (11)

• Adam: A Method for Stochastic Optimization
• Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
• Prototypical Networks for Few-shot Learning
• Recurrent Neural Network Regularization
• Zero-Shot Learning Through Cross-Modal Transfer
• Zero-Shot Learning by Convex Combination of Semantic Embeddings
• Learning to reinforcement learn
• Incorporating Copying Mechanism in Sequence-to-Sequence Learning
• Recasting Gradient-Based Meta-Learning as Hierarchical Bayes
• Few-Shot Generalization Across Dialogue Tasks
• Adversarial Domain Adaptation for Variational Neural Language Generation in Dialogue Systems

Cited by in corpus (9)

• Language Models as Few-Shot Learner for Task-Oriented Dialogue Systems
• Recent Advances and Challenges in Task-oriented Dialog System
• MALA: Cross-Domain Dialogue Generation with Action Learning
• Prototypical Q Networks for Automatic Conversational Diagnosis and Few-Shot New Disease Adaption
• Meta Dialogue Policy Learning
• ST${}^2$: Small-data Text Style Transfer via Multi-task Meta-Learning
• How to Build User Simulators to Train RL-based Dialog Systems
• Meta Learning to Rank for Sparsely Supervised Queries
• Resource Constrained Dialog Policy Learning via Differentiable Inductive Logic Programming