Proton: Probing Schema Linking Information from Pre-trained Language Models for Text-to-SQL Parsing. (arXiv:2206.14017v2 [cs.CL] UPDATED)

The importance of building text-to-SQL parsers which can be applied to new
databases has long been acknowledged, and a critical step to achieve this goal
is schema linking, i.e., properly recognizing mentions of unseen columns or
tables when generating SQLs. In this work, we propose a novel framework to
elicit relational structures from large-scale pre-trained language models
(PLMs) via a probing procedure based on Poincar\'e distance metric, and use the
induced relations to augment current graph-based parsers for better schema …

arxiv information language language models parsing proton schema sql text