Abstract
A natural language interface (NLI) to structured query is intriguing due toits wide industrial applications and high economical values. In this work, wetackle the problem of domain adaptation for NLI with limited data on targetdomain. Two important approaches are considered: (a) effectivegeneral-knowledge-learning on source domain semantic parsing, and (b) dataaugmentation on target domain. We present a Structured Query Inference Network(SQIN) to enhance learning for domain adaptation, by separating schemainformation from NL and decoding SQL in a more structural-aware manner; we alsopropose a GAN-based augmentation technique (AugmentGAN) to mitigate the issueof lacking target domain data. We report solid results on GeoQuery, Overnight,and WikiSQL to demonstrate state-of-the-art performances for both in-domain anddomain-transfer tasks.