🤖 AI Summary
Text-to-SQL models suffer from strong dependence on condition-column data distributions and human annotations in real-world settings, coupled with poor generalization in semantic parsing of WHERE clauses. To address these issues, this paper proposes a condition-augmentation method grounded in interpretability analysis and execution-guided learning. Our approach employs three core mechanisms—filtering-and-adjustment, logical-association refinement, and multi-model fusion—operating without additional labeled data, thereby substantially reducing model sensitivity to the value distribution of condition columns. Evaluated on the WikiSQL single-table benchmark, the method achieves significant improvements in overall execution accuracy. Notably, it sets a new state-of-the-art (SOTA) on the subtask of condition-value prediction within WHERE clauses, demonstrating enhanced fundamental semantic understanding and superior cross-instance generalization capability.
📝 Abstract
To elevate the foundational capabilities and generalization prowess of the text-to-SQL model in real-world applications, we integrate model interpretability analysis with execution-guided strategy for semantic parsing of WHERE clauses in SQL queries. Furthermore, we augment this approach with filtering adjustments, logical correlation refinements, and model fusion, culminating in the design of the CESQL model that facilitates conditional enhancement. Our model excels on the WikiSQL dataset, which is emblematic of single-table database query tasks, markedly boosting the accuracy of prediction outcomes. When predicting conditional values in WHERE clauses, we have not only minimized our dependence on data within the condition columns of tables but also circumvented the impact of manually labeled training data. Our hope is that this endeavor to enhance accuracy in processing basic database queries will offer fresh perspectives for research into handling complex queries and scenarios featuring irregular data in real-world database environments.