Calibrated Interpretation: Confidence Estimation in Semantic Parsing

TL;DR

This paper examines the calibration of sequence generation models in semantic parsing, highlighting variability across models and datasets, and provides tools and datasets to improve safety and confidence estimation in real-world applications.

Contribution

It analyzes calibration issues in semantic parsing models, introduces new challenge datasets, and releases a library for calibration metrics to enhance safety evaluations.

Findings

Calibration varies across models and datasets.

Factors influencing calibration errors are identified.

New challenge splits and a calibration metrics library are released.

Abstract

Sequence generation models are increasingly being used to translate natural language into programs, i.e. to perform executable semantic parsing. The fact that semantic parsing aims to predict programs that can lead to executed actions in the real world motivates developing safe systems. This in turn makes measuring calibration -- a central component to safety -- particularly important. We investigate the calibration of popular generation models across four popular semantic parsing datasets, finding that it varies across models and datasets. We then analyze factors associated with calibration error and release new confidence-based challenge splits of two parsing datasets. To facilitate the inclusion of calibration in semantic parsing evaluations, we release a library for computing calibration metrics.