Representing Schema Structure with Graph Neural Networks for Text-to-SQL Parsing
This addresses the challenge of parsing natural language to SQL queries for complex databases in the Spider dataset, representing a strong specific gain rather than an incremental improvement.
The paper tackled the problem of text-to-SQL parsing by incorporating database schema structure using a graph neural network, resulting in an accuracy improvement from 33.8% to 39.4%, which is significantly above the previous state-of-the-art of 19.7%.
Research on parsing language to SQL has largely ignored the structure of the database (DB) schema, either because the DB was very simple, or because it was observed at both training and test time. In Spider, a recently-released text-to-SQL dataset, new and complex DBs are given at test time, and so the structure of the DB schema can inform the predicted SQL query. In this paper, we present an encoder-decoder semantic parser, where the structure of the DB schema is encoded with a graph neural network, and this representation is later used at both encoding and decoding time. Evaluation shows that encoding the schema structure improves our parser accuracy from 33.8% to 39.4%, dramatically above the current state of the art, which is at 19.7%.