Error Detection for Text-to-SQL Semantic Parsing

TL;DR

This paper introduces a parser-independent error detection model for text-to-SQL semantic parsing that leverages code language models and graph neural networks to improve error identification and overall parser performance.

Contribution

We propose a novel error detection approach that is independent of specific parsers, utilizing language models and graph neural networks trained on realistic errors for better generalization.

Findings

Outperforms parser-dependent uncertainty metrics in error detection

Enhances the performance of various text-to-SQL parsers

Improves usability and trustworthiness of semantic parsing systems

Abstract

Despite remarkable progress in text-to-SQL semantic parsing in recent years, the performance of existing parsers is still far from perfect. Specifically, modern text-to-SQL parsers based on deep learning are often over-confident, thus casting doubt on their trustworthiness when deployed for real use. In this paper, we propose a parser-independent error detection model for text-to-SQL semantic parsing. Using a language model of code as its bedrock, we enhance our error detection model with graph neural networks that learn structural features of both natural language questions and SQL queries. We train our model on realistic parsing errors collected from a cross-domain setting, which leads to stronger generalization ability. Experiments with three strong text-to-SQL parsers featuring different decoding mechanisms show that our approach outperforms parser-dependent uncertainty metrics. Our model could also effectively improve the performance and usability of text-to-SQL semantic parsers regardless of their architectures. (Our implementation is available at https://github.com/OSU-NLP-Group/Text2SQL-Error-Detection)