Weakly Supervised Text-to-SQL Parsing through Question Decomposition

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Wolfson, Tomer, Deutch, Daniel, Berant, Jonathan
Formato: Preprint
Publicado: 2021
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866913455441707008
author Wolfson, Tomer
Deutch, Daniel
Berant, Jonathan
author_facet Wolfson, Tomer
Deutch, Daniel
Berant, Jonathan
contents Text-to-SQL parsers are crucial in enabling non-experts to effortlessly query relational data. Training such parsers, by contrast, generally requires expertise in annotating natural language (NL) utterances with corresponding SQL queries. In this work, we propose a weak supervision approach for training text-to-SQL parsers. We take advantage of the recently proposed question meaning representation called QDMR, an intermediate between NL and formal query languages. Given questions, their QDMR structures (annotated by non-experts or automatically predicted), and the answers, we are able to automatically synthesize SQL queries that are used to train text-to-SQL models. We test our approach by experimenting on five benchmark datasets. Our results show that the weakly supervised models perform competitively with those trained on annotated NL-SQL data. Overall, we effectively train text-to-SQL parsers, while using zero SQL annotations.
format Preprint
id arxiv_https___arxiv_org_abs_2112_06311
institution arXiv
publishDate 2021
record_format arxiv
spellingShingle Weakly Supervised Text-to-SQL Parsing through Question Decomposition
Wolfson, Tomer
Deutch, Daniel
Berant, Jonathan
Computation and Language
Artificial Intelligence
Databases
Text-to-SQL parsers are crucial in enabling non-experts to effortlessly query relational data. Training such parsers, by contrast, generally requires expertise in annotating natural language (NL) utterances with corresponding SQL queries. In this work, we propose a weak supervision approach for training text-to-SQL parsers. We take advantage of the recently proposed question meaning representation called QDMR, an intermediate between NL and formal query languages. Given questions, their QDMR structures (annotated by non-experts or automatically predicted), and the answers, we are able to automatically synthesize SQL queries that are used to train text-to-SQL models. We test our approach by experimenting on five benchmark datasets. Our results show that the weakly supervised models perform competitively with those trained on annotated NL-SQL data. Overall, we effectively train text-to-SQL parsers, while using zero SQL annotations.
title Weakly Supervised Text-to-SQL Parsing through Question Decomposition
topic Computation and Language
Artificial Intelligence
Databases
url https://arxiv.org/abs/2112.06311