Improving Retrieval-augmented Text-to-SQL with AST-based Ranking and Schema Pruning

TL;DR

This paper introduces ASTReS, a retrieval-augmented approach for Text-to-SQL parsing that uses AST-based ranking and schema pruning to improve performance on monolingual and cross-lingual benchmarks.

Contribution

The paper presents ASTReS, a novel method combining dynamic retrieval, AST-based example selection, and a lightweight semantic parser for efficient, improved Text-to-SQL translation.

Findings

ASTReS outperforms state-of-the-art baselines on multiple benchmarks.

Schema pruning and AST-based ranking significantly enhance retrieval accuracy.

A small, efficient model effectively supports parallel schema processing.

Abstract

We focus on Text-to-SQL semantic parsing from the perspective of retrieval-augmented generation. Motivated by challenges related to the size of commercial database schemata and the deployability of business intelligence solutions, we propose $\text{ASTReS}$ that dynamically retrieves input database information and uses abstract syntax trees to select few-shot examples for in-context learning. Furthermore, we investigate the extent to which an in-parallel semantic parser can be leveraged for generating approximated versions of the expected SQL queries, to support our retrieval. We take this approach to the extreme--we adapt a model consisting of less than $500$M parameters, to act as an extremely efficient approximator, enhancing it with the ability to process schemata in a parallelised manner. We apply $\text{ASTReS}$ to monolingual and cross-lingual benchmarks for semantic parsing, showing improvements over state-of-the-art baselines. Comprehensive experiments highlight the contribution of modules involved in this retrieval-augmented generation setting, revealing interesting directions for future work.