A metadata-based database foreign key relationship identification method

By collecting database metadata and combining explicit primary key and semantic primary key recognition methods, and utilizing multi-dimensional feature fusion to calculate structural dependency index, the accuracy and stability issues of foreign key relationship recognition in complex databases are solved, achieving efficient foreign key relationship recognition and database structure optimization.

CN122173496APending Publication Date: 2026-06-09HUNAN UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
HUNAN UNIV
Filing Date
2026-03-26
Publication Date
2026-06-09

AI Technical Summary

Technical Problem

Existing methods for identifying foreign key relationships lack accuracy and stability in complex databases, making it difficult to achieve accurate identification in situations where complete schema information is lacking, data noise is present, and naming conventions are inconsistent.

Method used

By collecting metadata information from the database, combining explicit primary key and semantic primary key identification methods, and utilizing multi-dimensional feature fusion and quantitative evaluation, structural dependency indexes are calculated to identify foreign key relationships between database tables.

Benefits of technology

It achieves high-accuracy foreign key relationship identification in complex databases, provides reliable structural information support, and helps to improve database structure and data quality.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122173496A_ABST
    Figure CN122173496A_ABST
Patent Text Reader

Abstract

This invention discloses a method for identifying foreign key relationships in databases based on metadata, relating to database technology, data processing, and natural language processing. The method first extracts database metadata and statistically analyzes core field features, generating a standardized metadata set through standardized preprocessing. Then, it parses this set to filter explicit primary keys, determines semantic primary keys based on metadata statistical features and semantic rules, and integrates these to generate a global primary key set. Subsequently, it encodes string features in the metadata set to generate candidate foreign key field pairs, and calculates the structural dependency degree (SDS) by fusing three metadata-derived features: value range coverage, primary key uniqueness coefficient, and reference dependency rate. Finally, it verifies and filters valid and suspected foreign key relationships based on a preset threshold, generating a standardized foreign key relationship result set. This invention does not rely on field naming, is robust to data noise, accurately identifies database foreign key relationships, and achieves full-process traceability and verifiability based on metadata, providing reliable structural support for data governance and intelligent database applications.
Need to check novelty before this filing date? Find Prior Art