UniSAr: A Unified Structure-Aware Autoregressive Language Model for Text-to-SQL
This work addresses the need for a versatile text-to-SQL parser that can handle multiple settings without redesign, offering a non-invasive solution that can integrate with other advances in the field.
The paper tackles the problem of text-to-SQL semantic parsing across diverse settings like multi-domain or multi-turn queries by introducing UniSAr, a unified structure-aware autoregressive language model that achieves highly comparable or better performance than specialized models on seven datasets.
Existing text-to-SQL semantic parsers are typically designed for particular settings such as handling queries that span multiple tables, domains or turns which makes them ineffective when applied to different settings. We present UniSAr (Unified Structure-Aware Autoregressive Language Model), which benefits from directly using an off-the-shelf language model architecture and demonstrates consistently high performance under different settings. Specifically, UniSAr extends existing autoregressive language models to incorporate three non-invasive extensions to make them structure-aware: (1) adding structure mark to encode database schema, conversation context, and their relationships; (2) constrained decoding to decode well structured SQL for a given database schema; and (3) SQL completion to complete potential missing JOIN relationships in SQL based on database schema. On seven well-known text-to-SQL datasets covering multi-domain, multi-table and multi-turn, UniSAr demonstrates highly comparable or better performance to the most advanced specifically-designed text-to-SQL models. Importantly, our UniSAr is non-invasive, such that other core model advances in text-to-SQL can also adopt our extensions to further enhance performance.