Supply Chain SQL Generation Hard-Case Dataset

#SQL generation #Text-to-SQL #SQL error diagnosis #Schema understanding #Supply chain analytics #Procurement reporting #Inventory analysis #Order fulfillment

7749
0.5GB
JSON
CATL
MOBIUSI INC

Updated:2026-11-03

AI Analysis & Value Prop

This dataset contains natural-language questions and SQL examples that are prone to errors in supply chain database queries, covering pitfalls such as join fan-out, missing-value handling, temporal boundaries, and mismatched business grains. Each example pairs a database schema with reference SQL and optional business context to help models interpret intent and generate queries that respect both the schema and business rules. It supports supervised SQL Agent training, text-to-SQL evaluation, and query error analysis.

Dataset Insights

Sample Examples

{}

Technical Specifications

Field	Type	Description
file_name	string	File name
query_text	string	The original business question that the SQL query is expected to answer.
reference_sql	string	The vetted SQL query that correctly answers the natural-language question against the supplied schema.
database_schema	string	The source database schema, including relevant tables, columns, and relationships, represented as SQL DDL or equivalent schema text.
business_context	string	Optional domain definitions, metric rules, or other business assumptions needed to interpret the query correctly.
join_fanout_risk	boolean	Indicates whether joins across tables could duplicate rows and inflate counts or aggregates if handled incorrectly.
failure_mechanism	string	A concise explanation of how a plausible but incorrect query could produce an erroneous result.
business_grain_note	string	An explanation of the intended business entity or aggregation grain and any mismatch that must be resolved.
hard_case_categories	array	Derived labels for the semantic or structural pitfalls addressed by the example.
missing_value_handling	string	Guidance on how null or otherwise missing values should be treated for the query, or an indication that no special handling is needed.
temporal_boundary_rule	string	Guidance on relevant date or time filters, including boundary inclusivity and interval interpretation, or an indication that no temporal boundary applies.

Compliance Statement

Authorization Type	Proprietary - Commercial AI Training License (No Redistribution)
Commercial Use	Requires exclusive subscription or authorization contract (monthly or per-invocation charging)
Privacy and Anonymization	No PII, no real company names, simulated scenarios follow industry standards
Compliance System	Compliant with China's Data Security Law / EU GDPR / supports enterprise data access logs

Frequently Asked Questions

What does the dataset contain?: The examples focus on supply chain database queries and include natural-language questions, relevant database schemas, and reference SQL, with optional business context and explanations of the pitfalls. This supports query interpretation grounded in both schema and business semantics.

Which SQL challenges are covered?: The dataset focuses on join fan-out that can inflate counts, missing-value handling, date or time interval boundaries, and mismatches between business entities and aggregation grains. Example categories and notes describe the specific issue addressed.

How can I obtain the dataset?: Check the product page for available data formats, delivery methods, and access instructions. The actual contents and delivery details are subject to the product listing and accompanying documentation.

What are the licensing terms and intended uses?: The dataset is suitable for supervised SQL Agent training, text-to-SQL evaluation, query error analysis, and related development work. Please consult the product page or accompanying license terms for commercial use, redistribution, and other permitted uses.

Can't find the data you need?

Post a request and let data providers reach out to you.

Get this Dataset

Verified for Enterprise Use

Cite this Work

@dataset{Mobiusi2026,
  title={Supply Chain SQL Generation Hard-Case Dataset},
  author={MOBIUSI INC},
  year={2026},
  url={https://www.mobiusi.com/datasets/d60491592266c02c01c06c5b0458d1cd},
  urldate={2026-10-22},
  keywords={supply chain SQL dataset,SQL generation training data,text-to-SQL examples,SQL hard cases,SQL Agent training,supply chain database queries,multi-table join SQL,SQL error analysis},
  version={1.0}
}

Using this in research? Please cite us.

placeholder

Scene

Domain

Modality

Task