mlsanityData to 7 Oct 2026

Schema-Aware Localisation (SAL): Live Schema Grounding and Hallucination Validation for Oracle NL2SQL

arXiv 2607.22572 ↗First posted 3 Jun 2026cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Large language models can generate fluent SQL from natural language, but on real enterprise Oracle databases they frequently fail at execution time: columns and aliases are hallucinated and dialect-specific syntax is missed, leading to ORA-00904 invalid-identifier errors. In this setting, failures are primarily due to missing schema grounding: the model cannot know which tables and columns actually exist. This paper introduces Schema-Aware Localisation (SAL), a lightweight middleware layer for Oracle NL2SQL that requires no model retraining. SAL queries Oracle's USER_TAB_COLUMNS catalog to build a live schema map, selects a relevant table subset for each question (falling back to the full schema for multi-table queries), and injects this ground-truth context into the LLM prompt. Generated SQL is then checked by the Hallucination Index (Hidx), which validates every alias.column reference against the live catalog, automatically rewrites predictable prefix errors, and otherwise triggers a structured retry with itemised corrections. We evaluate SAL on 500 TPC-H natural language questions executed against a live Oracle Autonomous Database 23c instance using GPT-4o-mini. Without any schema grounding, execution-grounded truth (EGT; executes and matches the reference result set) is 2.2% (12/500). A hand-written static schema hint brings EGT to 62.0%. SAL, with no manual schema curation, achieves 62.6% EGT (96% simple, 95% medium, 40.7% complex) while reducing execution failures from 97.6% to 2.6%.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 21 of its 51 references that are in the map. First posted on arXiv 3 Jun 2026.

Builds on

21 of 51 references.

References · 21

  1. 2019Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformercited by 14,491
  2. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,302
  3. 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,160
  4. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,355
  5. 2023Toolformer: Language Models Can Teach Themselves to Use Toolscited by 4,072
  6. 2023Lost in the Middle: How Language Models Use Long Contextscited by 3,383
  7. 2022Survey of Hallucination in Natural Language Generationcited by 2,056
  8. 2023Is Your Code Generated by ChatGPT Really Correct? Rigorous Evaluation of Large Language Models for Code Generationcited by 1,332
  9. 2018Spider: A Large-Scale Human-Labeled Dataset for Complex and Cross-Domain Semantic Parsing and Text-to-SQL Taskcited by 1,120
  10. 2020On Faithfulness and Factuality in Abstractive Summarizationcited by 895
  11. 2017Seq2SQL: Generating Structured Queries from Natural Language using Reinforcement Learningcited by 679
  12. 2023DIN-SQL: Decomposed In-Context Learning of Text-to-SQL with Self-Correctioncited by 465
9 more
  1. 2019RAT-SQL: Relation-Aware Schema Encoding and Linking for Text-to-SQL Parserscited by 420
  2. 2023Text-to-SQL Empowered by Large Language Models: A Benchmark Evaluationcited by 378
  3. 2020GraPPa: Grammar-Augmented Pre-Training for Table Semantic Parsingcited by 171
  4. 2023C3: Zero-shot Text-to-SQL with ChatGPTcited by 159
  5. 2016Language to Logical Form with Neural Attentioncited by 158
  6. 2017SQLNet: Generating Structured Queries From Natural Language Without Reinforcement Learningcited by 135
  7. 2021LGESQL: Line Graph Enhanced Text-to-SQL Model with Mixed Local and Non-Local Relationscited by 119
  8. 2018TypeSQL: Knowledge-based Type-Aware Neural Text-to-SQL Generationcited by 81
  9. 2018Robust Text-to-SQL Generation with Execution-Guided Decodingcited by 51

Built on by

Not cited yet.