Multi-Constraint Fashion Product Catalog Retrieval Benchmark

#Document Retrieval #Query Constraint Matching #Query-Product Relevance Ranking #Query-Document Matching #Fashion Catalog Search #Premium Retail Product Discovery #E-commerce Product Retrieval #Designer Product Filtering

4260
0.5GB
JSON
CATL
MOBIUSI INC

Updated:2026-11-03

AI Analysis & Value Prop

This benchmark targets complex retrieval needs in designer and premium fashion retail catalogs. It pairs natural-language queries containing one or more constraints, such as category, material, silhouette, color, or use occasion, with catalog products for evaluation. Structured query constraints, constraint-match summaries, relevance grades, and supporting explanations make it suitable for assessing query understanding, product retrieval, query-document matching, and relevance ranking; the total record count is not specified in the available information.

Dataset Insights

Sample Examples

{}

Technical Specifications

Field	Type	Description
file_name	string	File name
query_text	string	The original natural-language query, which may specify one or more fashion product constraints.
catalog_product	object	The catalog product record evaluated for relevance to the query.
relevance_grade	integer	An ordinal relevance judgment: 0 means not relevant, 1 weakly relevant, 2 partially relevant, and 3 highly relevant.
relevance_reason	string	An explanation for the relevance grade, based on the query constraints and product attributes.
query_constraints	object	Structured constraints extracted from the original query to support analysis and product matching.
matched_constraints	string	A concise summary of the query constraints satisfied or contradicted by the product.

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 benchmark contain?: Each record includes a natural-language product query, a catalog product, parsed query constraints, a summary of which constraints match, a relevance grade from 0 to 3, and a supporting explanation. Query constraints may cover category, material, silhouette, color, and use occasion.

Are the dataset size and annotation scheme specified?: The available product information does not specify the total number of records. Each record includes structured constraints and a relevance grade: 0 means not relevant, 1 weakly relevant, 2 partially relevant, and 3 highly relevant, with a supporting rationale.

How can I obtain or use the benchmark?: Follow the download, delivery, or access instructions on the current product page. The benchmark is suited to training and evaluating models for query-to-product retrieval, complex constraint matching, and relevance ranking.

What are the licensing terms?: The applicable license and permitted uses depend on the terms provided with this product. Check those terms before commercial deployment, redistribution, or public release of evaluation results.

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={Multi-Constraint Fashion Product Catalog Retrieval Benchmark},
  author={MOBIUSI INC},
  year={2026},
  url={https://www.mobiusi.com/datasets/a022b393a5f8be288863a1c9b797de66?dataset_scene_cate_type=4},
  urldate={2026-10-22},
  keywords={fashion product retrieval benchmark,product catalog search dataset,multi-constraint product queries,fashion e-commerce retrieval,query-document matching,relevance-labeled products,premium fashion retail,product search evaluation},
  version={1.0}
}

Using this in research? Please cite us.

placeholder

Scene

Domain

Modality

Task