mlsanityData to 14 Oct 2026

Schema-First Retrieval: Embedding Catalogs for Natural Language Analytics

arXiv 2606.28387 ↗First posted 23 Jun 2026cs.IR cs.AIcited by 1

Not on the map. Why?

Abstract (arXiv)

Enterprise text-to-SQL systems often fail before SQL is generated: the model receives the wrong schema context. Modern warehouses contain thousands of tables, abbreviated columns, informal metrics, hidden join conventions, and permission boundaries that are not captured by raw table names. We introduce Schema-First Retrieval, a retrieval layer that embeds catalog metadata rather than warehouse rows. The system indexes five typed catalog objects, tables, columns, metrics, relationships, and query history, using object-specific text templates. At query time, it combines parallel vector search, lineage expansion, cross-encoder reranking, workload memory, and deterministic access-control gates before SQL generation. On CRUSH4SQL (1,534 questions), Schema-First Retrieval reaches 96.4% table recall@20 and cross-encoder reranking adds +11.1 points at column recall@10; against an equally-templated BM25 baseline, semantic retrieval is +32.8 points at table recall@5. On SEDE (857 questions), query history raises table recall@5 from 52.1% to 92.3%. On BIRD (96 questions), schema-first context reduces SQL execution errors from 15.6% to 6.2%, a 2.5x reduction. These results show that catalog selection is a first-class retrieval problem for natural language analytics, not a prompt formatting detail.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 17 of its 25 references that are in the map. First posted on arXiv 23 Jun 2026.

Builds on

17 of 25 references.

References · 17

  1. 2023DIN-SQL: Decomposed In-Context Learning of Text-to-SQL with Self-Correctioncited by 477
  2. 2023Text-to-SQL Empowered by Large Language Models: A Benchmark Evaluationcited by 385
  3. 2024Spider 2.0: Evaluating Language Models on Real-World Enterprise Text-to-SQL Workflowscited by 232
  4. 2023MAC-SQL: A Multi-Agent Collaborative Framework for Text-to-SQLcited by 164
  5. 2021Text-to-SQL in the Wild: A Naturally-Occurring Dataset Based on Stack Exchange Datacited by 43
  6. 2025The Amazon Nova Family of Models: Technical Report and Model Cardcited by 41
  7. 2025LinkAlign: Scalable Schema Linking for Real-World Large-Scale Multi-Database Text-to-SQLcited by 29
  8. 2025Retrieval-Augmented Generation: A Comprehensive Survey of Architectures, Enhancements, and Robustness Frontierscited by 27
  9. 2025AutoLink: Autonomous Schema Exploration and Expansion for Scalable Schema Linking in Text-to-SQL at Scalecited by 17
  10. 2025Extractive Schema Linking for Text-to-SQLcited by 13
  11. 2024Solid-SQL: Enhanced Schema-linking based In-context Learning for Robust Text-to-SQLcited by 11
  12. 2025Permissioned LLMs: Enforcing Access Control in Large Language Modelscited by 7
5 more
  1. 2026Self-Healing Agentic Orchestrators for Reliable Tool-Augmented Large Language Model Systemscited by 7
  2. 2025RASL: Retrieval Augmented Schema Linking for Massive Database Text-to-SQLcited by 4
  3. 2025Rethinking Schema Linking: A Context-Aware Bidirectional Retrieval Approach for Text-to-SQLcited by 4
  4. 2025SchemaGraphSQL: Efficient Schema Linking with Pathfinding Graph Algorithms for Text-to-SQL on Large-Scale Databasescited by 3
  5. 2025Enterprise AI Must Enforce Participant-Aware Access Controlcited by 3

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026Grounded Optimization: A Layered Engineering Framework for Reducing LLM Hallucination in Automated Personal Document Rewritingcited by 0