Scaling Text2SQL via LLM-efficient Schema Filtering with Functional Dependency Graph Rerankers
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866915684285415424 |
|---|---|
| author | Hoang, Thanh Dat Nguyen, Thanh Tam Huynh, Thanh Trung Yin, Hongzhi Nguyen, Quoc Viet Hung |
| author_facet | Hoang, Thanh Dat Nguyen, Thanh Tam Huynh, Thanh Trung Yin, Hongzhi Nguyen, Quoc Viet Hung |
| contents | Most modern Text2SQL systems prompt large language models (LLMs) with entire schemas -- mostly column information -- alongside the user's question. While effective on small databases, this approach fails on real-world schemas that exceed LLM context limits, even for commercial models. The recent Spider 2.0 benchmark exemplifies this with hundreds of tables and tens of thousands of columns, where existing systems often break. Current mitigations either rely on costly multi-step prompting pipelines or filter columns by ranking them against user's question independently, ignoring inter-column structure. To scale existing systems, we introduce \toolname, an open-source, LLM-efficient schema filtering framework that compacts Text2SQL prompts by (i) ranking columns with a query-aware LLM encoder enriched with values and metadata, (ii) reranking inter-connected columns via a lightweight graph transformer over functional dependencies, and (iii) selecting a connectivity-preserving sub-schema with a Steiner-tree heuristic. Experiments on real datasets show that \toolname achieves near-perfect recall and higher precision than CodeS, SchemaExP, Qwen rerankers, and embedding retrievers, while maintaining sub-second median latency and scaling to schemas with 23,000+ columns. Our source code is available at https://github.com/thanhdath/grast-sql. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_16083 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Scaling Text2SQL via LLM-efficient Schema Filtering with Functional Dependency Graph Rerankers Hoang, Thanh Dat Nguyen, Thanh Tam Huynh, Thanh Trung Yin, Hongzhi Nguyen, Quoc Viet Hung Databases Artificial Intelligence Human-Computer Interaction Machine Learning Most modern Text2SQL systems prompt large language models (LLMs) with entire schemas -- mostly column information -- alongside the user's question. While effective on small databases, this approach fails on real-world schemas that exceed LLM context limits, even for commercial models. The recent Spider 2.0 benchmark exemplifies this with hundreds of tables and tens of thousands of columns, where existing systems often break. Current mitigations either rely on costly multi-step prompting pipelines or filter columns by ranking them against user's question independently, ignoring inter-column structure. To scale existing systems, we introduce \toolname, an open-source, LLM-efficient schema filtering framework that compacts Text2SQL prompts by (i) ranking columns with a query-aware LLM encoder enriched with values and metadata, (ii) reranking inter-connected columns via a lightweight graph transformer over functional dependencies, and (iii) selecting a connectivity-preserving sub-schema with a Steiner-tree heuristic. Experiments on real datasets show that \toolname achieves near-perfect recall and higher precision than CodeS, SchemaExP, Qwen rerankers, and embedding retrievers, while maintaining sub-second median latency and scaling to schemas with 23,000+ columns. Our source code is available at https://github.com/thanhdath/grast-sql. |
| title | Scaling Text2SQL via LLM-efficient Schema Filtering with Functional Dependency Graph Rerankers |
| topic | Databases Artificial Intelligence Human-Computer Interaction Machine Learning |
| url | https://arxiv.org/abs/2512.16083 |