Macro Grammars and Holistic Triggering for Efficient Semantic Parsing

Yuchen Zhang, Panupong Pasupat, Percy Liang
7/25/2017
cs.CL

Official code: percyliang/sempre (842 stars, last commit 2023-05-01).

Abstract

To learn a semantic parser from denotations, a learning algorithm must search over a combinatorially large space of logical forms for ones consistent with the annotated denotations. We propose a new online learning algorithm that searches faster as training progresses. The two key ideas are using macro grammars to cache the abstract patterns of useful logical forms found thus far, and holistic triggering to efficiently retrieve the most relevant patterns based on sentence similarity. On the WikiTableQuestions dataset, we first expand the search space of an existing model to improve the state-of-the-art accuracy from 38.7% to 42.7%, and then use macro grammars and holistic triggering to achieve an 11x speedup and an accuracy of 43.7%.

View on arXivView PDF

Code Implementations(1)

percyliang/sempreOfficial100%

Semantic Parser with Execution

842293CSS, HTMLSep 18, 20133 years agoNOASSERTION

Similar Papers

Inferring Logical Forms From Denotations

86% match1 repo8422016

A Probabilistic Generative Grammar for Semantic Parsing

81% match2 repos302016

From Language to Programs: Bridging Reinforcement Learning and Maximum Marginal Likelihood

81% match2017

Semantic Parsing with Candidate Expressions for Knowledge Base Question Answering

80% match1 repo2024

Semantic Operator Prediction and Applications

80% match2023

TAPAS: Weakly Supervised Table Parsing via Pre-training

80% match1 repo1.2k2020

Cite this paper

@article{zhang2017macro,
  title  = {Macro Grammars and Holistic Triggering for Efficient Semantic Parsing},
  author = {Yuchen Zhang and Panupong Pasupat and Percy Liang},
  year   = {2017},
  eprint = {1707.07806},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1707.07806}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1707.07806.svg)](https://codeofpaper.com/paper/1707.07806)

Discussion