Skip to content

Can LLMs Normalize Databases? A Benchmark and Multi-Agent Framework for Schema Normalization

Sep 2026 · 0 citations · 37 references
Computer Science

TL;DR

A Database Normalization Benchmark (DNBENCH), comprising 3,275 samples for evaluating LLM-driven database normalization from 1NF to BCNF, and proposes Multi-Agent Reasoning for Schemas (MARS), which separates evidence extraction, violation diagnosis, and decomposition planning from schema generation and verification.

Abstract

Large Language Models (LLMs) are increasingly used to generate structured outputs, but their reliability remains unclear when those outputs must satisfy database-level constraints. We study this issue through database normalization, involving reasoning about functional dependencies, lossless join decompositions, and inter-table constraints. We introduce a Database Normalization Benchmark (DNBENCH), comprising 3,275 samples for evaluating LLM-driven database normalization from 1NF to BCNF. DNBENCH uses a three-axis protocol to measure semantic equivalence, structural accuracy, and logical validity. Across Single, Complex, and Real World levels, DNBENCH uncovers recurring failures in dependency inference, schema decomposition, and inter-table constraint reconstruction. We further propose Multi-Agent Reasoning for Schemas (MARS), which separates evidence extraction, violation diagnosis, and decomposition planning from schema generation and verification. MARS improves the DNB-SCORE by 82.0% over the single-prompt baseline. All artifacts will be released upon acceptance.

View source

Similar papers

#artificial intelligence Preprint Sep 2026

A Cost-Aware Agentic Architecture for NL-to-SQL over Nested Enterprise Schemas, with a New Benchmark

A cost-aware single- generation agentic architecture whose schema- selection, metadata-retrieval, and error-repair components are designed for the requirements this regime imposes, and is competitive with leading systems at a single-generation operating point.

Yoga Sri Varshan Varadharajan, Ajay Yadav, Ritesh Goru et al. · 0 citations
#artificial intelligence Preprint Sep 2026

SemVerBench: Benchmarking LLM Comprehension of Version-Constraint Resolution Semantics

SemVerBench is introduced, the first benchmark of LLM version-constraint resolution semantics across three ecosystems (npm, PEP 440, Cargo), and six frontier models are evaluated: Opus significantly outperforms all other models, and Sonnet outperforms the OpenAI models (McNemar).

Qi-Bai Chen, Ze-Ming Liu · 2 citations
Sep 2026

Focus on Evidence: Relational-Structure Enhances LLM Effectiveness in TableQA

EVA first extracts evidence-aware relational structures that associate question semantics with supporting table attributes that constraining sub-question generation within a grounded evidence space, and introduces relation-guided validation and integrates both DP and PyAgent reasoning to improve robustness across diffe...

Zhen Yang, Zi-Wei Du, Ming-Han Zhang et al. · 0 citations
Preprint Aug 2026

ACTS-SQL: Agentic and Critic-Oriented Tree-Structured SQL Correctness with Large Language Models

This work presents a training-free framework that formulates SQL correction as a plan-guided, tree-structured debugging process that mitigates error accumulation during iterative refinement and demonstrates the effectiveness and stability of the approach in real-world deployments.

Xinmei Huang, Jie Song, Peng Li et al. · 0 citations
#artificial intelligence Preprint Sep 2026

Surprising Effectiveness of Self-Demonstrations in Enhancing Schema-Ontology Mapping with LLMs

This paper presents a self-demonstration-driven approach that combines a neuro-symbolic task decomposition with a novel mechanism for automatically generating pattern-guided, dependency-aware demonstrations to address the integration challenge of heterogeneous relational databases into a centralized ontology.

Siddhesh Thombre, Manasi S. Patwardhan, Sunita Sarawagi · 0 citations
#artificial intelligence Preprint Sep 2026

KC-Bench: A Dynamic Interactive Benchmark for Evaluating Knowledge Conflicts in LLM Agents

KC-Bench is introduced, a controlled multi-turn benchmark for measuring model-level behavior across world-knowledge conflicts, input inconsistencies, and multi-source temporal conflicts, and provides a reproducible diagnostic for developing conflict-aware reasoning and execution safeguards.

Yaxing Lyu, Sheng-Jie Zhou, B. Toh et al. · 2 citations

Related blog posts

MIT News · Artificial Intelligence Nov 17, 2026

Learning the language of machine learning

By studying the computational foundations of language learning, Associate Professor Jacob Andreas builds intelligent systems that can communicate effectively with humans.

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.