Speak to your Parser: Interactive Text-to-SQL with Natural Language Feedback
This addresses the challenge of improving interactive natural language interfaces for databases, though it is incremental as it builds on existing semantic parsing methods.
The paper tackles the problem of correcting semantic parsing errors in text-to-SQL systems by allowing users to provide natural language feedback, resulting in a dataset (SPLASH) and models that improve parsing accuracy, with a best model achieving 25.1% correction accuracy compared to an estimated human accuracy of 81.5%.
We study the task of semantic parse correction with natural language feedback. Given a natural language utterance, most semantic parsing systems pose the problem as one-shot translation where the utterance is mapped to a corresponding logical form. In this paper, we investigate a more interactive scenario where humans can further interact with the system by providing free-form natural language feedback to correct the system when it generates an inaccurate interpretation of an initial utterance. We focus on natural language to SQL systems and construct, SPLASH, a dataset of utterances, incorrect SQL interpretations and the corresponding natural language feedback. We compare various reference models for the correction task and show that incorporating such a rich form of feedback can significantly improve the overall semantic parsing accuracy while retaining the flexibility of natural language interaction. While we estimated human correction accuracy is 81.5%, our best model achieves only 25.1%, which leaves a large gap for improvement in future research. SPLASH is publicly available at https://aka.ms/Splash_dataset.