Structured Case-based Reasoning for Inference-time Adaptation of Text-to-SQL parsers

TL;DR

This paper introduces StructCBR, a structured case-based reasoning method that improves inference-time adaptation of Text-to-SQL parsers by leveraging subtree-level similarities in logical forms, outperforming prior methods.

Contribution

It presents the first inference-time adaptation approach for Text-to-SQL models using trainable structured similarity between subqueries, enhancing adaptation to unseen schemas.

Findings

Consistent performance improvements across five databases.

First to explore inference-time adaptation for Text-to-SQL.

Utilizes subtree-level similarity for better decoder guidance.

Abstract

Inference-time adaptation methods for semantic parsing are useful for leveraging examples from newly-observed domains without repeated fine-tuning. Existing approaches typically bias the decoder by simply concatenating input-output example pairs (cases) from the new domain at the encoder's input in a Seq-to-Seq model. Such methods cannot adequately leverage the structure of logical forms in the case examples. We propose StructCBR, a structured case-based reasoning approach, which leverages subtree-level similarity between logical forms of cases and candidate outputs, resulting in better decoder decisions. For the task of adapting Text-to-SQL models to unseen schemas, we show that exploiting case examples in a structured manner via StructCBR offers consistent performance improvements over prior inference-time adaptation methods across five different databases. To the best of our knowledge, we are the first to attempt inference-time adaptation of Text-to-SQL models, and harness trainable structured similarity between subqueries.