Bridging Global Intent with Local Details: A Hierarchical Representation Approach for Semantic Validation in Text-to-SQL

TL;DR

This paper introduces HEROSQL, a hierarchical SQL representation method combining global intent and local details, employing neural networks and data augmentation to improve semantic validation accuracy in Text-to-SQL systems.

Contribution

HEROSQL is a novel hierarchical approach that integrates logical plans and syntax trees with neural message passing and data augmentation for enhanced semantic validation.

Findings

Outperforms state-of-the-art in semantic inconsistency detection

Achieves 9.40% improvement in AUPRC

Achieves 12.35% improvement in AUROC

Abstract

Text-to-SQL translates natural language questions into SQL statements grounded in a target database schema. Ensuring the reliability and executability of such systems requires validating generated SQL, but most existing approaches focus only on syntactic correctness, with few addressing semantic validation (detecting misalignments between questions and SQL). As a consequence, effective semantic validation still faces two key challenges: capturing both global user intent and SQL structural details, and constructing high-quality fine-grained sub-SQL annotations. To tackle these, we introduce HEROSQL, a hierarchical SQL representation approach that integrates global intent (via Logical Plans, LPs) and local details (via Abstract Syntax Trees, ASTs). To enable better information propagation, we employ a Nested Message Passing Neural Network (NMPNN) to capture inherent relational information in SQL and aggregate schema-guided semantics across LPs and ASTs. Additionally, to generate high-quality negative samples, we propose an AST-driven sub-SQL augmentation strategy, supporting robust optimization of fine-grained semantic inconsistencies. Extensive experiments conducted on Text-to-SQL validation benchmarks (both in-domain and out-of-domain settings) demonstrate that our approach outperforms existing state-of-the-art methods, achieving an average 9.40% improvement of AUPRC and 12.35% of AUROC in identifying semantic inconsistencies. It excels at detecting fine-grained semantic errors, provides large language models with more granular feedback, and ultimately enhances the reliability and interpretability of data querying platforms.