One Semantic Parser to Parse Them All: Sequence to Sequence Multi-Task Learning on Semantic Parsing Datasets

Marco Damonte, Emilio Monti


Abstract
Semantic parsers map natural language utterances to meaning representations. The lack of a single standard for meaning representations led to the creation of a plethora of semantic parsing datasets. To unify different datasets and train a single model for them, we investigate the use of Multi-Task Learning (MTL) architectures. We experiment with five datasets (Geoquery, NLMaps, TOP, Overnight, AMR). We find that an MTL architecture that shares the entire network across datasets yields competitive or better parsing accuracies than the single-task baselines, while reducing the total number of parameters by 68%. We further provide evidence that MTL has also better compositional generalization than single-task models. We also present a comparison of task sampling methods and propose a competitive alternative to widespread proportional sampling strategies.
Anthology ID:
2021.starsem-1.16
Volume:
Proceedings of *SEM 2021: The Tenth Joint Conference on Lexical and Computational Semantics
Month:
August
Year:
2021
Address:
Online
Editors:
Lun-Wei Ku, Vivi Nastase, Ivan Vulić
Venue:
*SEM
SIG:
SIGLEX
Publisher:
Association for Computational Linguistics
Note:
Pages:
173–184
Language:
URL:
https://aclanthology.org/2021.starsem-1.16
DOI:
10.18653/v1/2021.starsem-1.16
Bibkey:
Cite (ACL):
Marco Damonte and Emilio Monti. 2021. One Semantic Parser to Parse Them All: Sequence to Sequence Multi-Task Learning on Semantic Parsing Datasets. In Proceedings of *SEM 2021: The Tenth Joint Conference on Lexical and Computational Semantics, pages 173–184, Online. Association for Computational Linguistics.
Cite (Informal):
One Semantic Parser to Parse Them All: Sequence to Sequence Multi-Task Learning on Semantic Parsing Datasets (Damonte & Monti, *SEM 2021)
Copy Citation:
PDF:
https://aclanthology.org/2021.starsem-1.16.pdf
Data
CFQ