Coarse-to-Fine Decoding for Neural Semantic Parsing

TL;DR

This paper introduces a two-stage neural semantic parsing method that first creates a rough sketch of meaning and then refines it, leading to improved performance across diverse datasets.

Contribution

A novel structure-aware coarse-to-fine decoding architecture for neural semantic parsing that enhances accuracy with simple decoders.

Findings

Consistent performance improvements across four datasets.

Achieves competitive results with simple decoders.

Effective decomposition of parsing into sketching and filling stages.

Abstract

Semantic parsing aims at mapping natural language utterances into structured meaning representations. In this work, we propose a structure-aware neural architecture which decomposes the semantic parsing process into two stages. Given an input utterance, we first generate a rough sketch of its meaning, where low-level information (such as variable names and arguments) is glossed over. Then, we fill in missing details by taking into account the natural language input and the sketch itself. Experimental results on four datasets characteristic of different domains and meaning representations show that our approach consistently improves performance, achieving competitive results despite the use of relatively simple decoders.