Knowledge Base Construction for Knowledge-Augmented Text-to-SQL

2025-05-28

AI Summary

Text-to-SQL faces challenges in domain knowledge coverage and generalization to unseen databases and diverse queries. To address this, we propose the first universal, transferable paradigm for constructing structured knowledge bases for Text-to-SQL, integrating database schemas, natural language questions, and semantically grounded associations. Our approach automatically extracts and organizes such knowledge—enabling zero-shot adaptation to novel databases—within a retrieval-augmented generation (RAG) framework, eliminating reliance on manual annotation or one-time generated knowledge. Evaluated across multiple standard benchmarks—including cross-domain and non-overlapping database settings—our method achieves an average 8.2%+ improvement in SQL execution accuracy over state-of-the-art baselines. This work establishes a reusable, cross-database, cross-domain knowledge infrastructure for Text-to-SQL, advancing robustness and scalability beyond task-specific fine-tuning.

Technology Category

• Natural Language Processing: Sentence-level Semantics, Textual Inference, etc.
• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval

Application Category

• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Text-to-SQL aims to translate natural language queries into SQL statements, which is practical as it enables anyone to easily retrieve the desired information from databases. Recently, many existing approaches tackle this problem with Large Language Models (LLMs), leveraging their strong capability in understanding user queries and generating corresponding SQL code. Yet, the parametric knowledge in LLMs might be limited to covering all the diverse and domain-specific queries that require grounding in various database schemas, which makes generated SQLs less accurate oftentimes. To tackle this, we propose constructing the knowledge base for text-to-SQL, a foundational source of knowledge, from which we retrieve and generate the necessary knowledge for given queries. In particular, unlike existing approaches that either manually annotate knowledge or generate only a few pieces of knowledge for each query, our knowledge base is comprehensive, which is constructed based on a combination of all the available questions and their associated database schemas along with their relevant knowledge, and can be reused for unseen databases from different datasets and domains. We validate our approach on multiple text-to-SQL datasets, considering both the overlapping and non-overlapping database scenarios, where it outperforms relevant baselines substantially.

Problem

• Enhancing Text-to-SQL accuracy with domain-specific knowledge
• Overcoming LLM limitations in diverse database queries
• Constructing reusable knowledge base for unseen databases

Innovation

• Constructing a knowledge base for Text-to-SQL
• Retrieving necessary knowledge from the base
• Reusing knowledge across diverse databases