RAT-SQL: Relation-Aware Schema Encoding and Linking for Text-to-SQL Parsers

TL;DR

This paper introduces RAT-SQL, a relation-aware schema encoding and linking framework that significantly improves text-to-SQL parsing accuracy, especially on unseen schemas, by enhancing schema representation and alignment modeling.

Contribution

The paper proposes a unified relation-aware self-attention mechanism for schema encoding and linking, achieving state-of-the-art results on the Spider dataset.

Findings

Achieves 57.2% exact match accuracy on Spider, surpassing previous best by 8.7%.

With BERT, reaches 65.6% accuracy, setting new state-of-the-art.

Qualitative improvements in schema linking and alignment understanding.

Abstract

When translating natural language questions into SQL queries to answer questions from a database, contemporary semantic parsing models struggle to generalize to unseen database schemas. The generalization challenge lies in (a) encoding the database relations in an accessible way for the semantic parser, and (b) modeling alignment between database columns and their mentions in a given query. We present a unified framework, based on the relation-aware self-attention mechanism, to address schema encoding, schema linking, and feature representation within a text-to-SQL encoder. On the challenging Spider dataset this framework boosts the exact match accuracy to 57.2%, surpassing its best counterparts by 8.7% absolute improvement. Further augmented with BERT, it achieves the new state-of-the-art performance of 65.6% on the Spider leaderboard. In addition, we observe qualitative improvements in the model's understanding of schema linking and alignment. Our implementation will be open-sourced at https://github.com/Microsoft/rat-sql.