AI-enhanced data bloodline analysis method and system
By using AI-enhanced parsing to generate candidate field mapping relationships and candidate lineage dependencies, the problem of missing field mapping relationships and incomplete flow paths in data lineage analysis in existing technologies is solved, and stable generation of fused lineage data and graph construction are achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- GUANGDONG TECSUN SCIENCE & TECHNOLOGY CO LTD
- Filing Date
- 2026-06-05
- Publication Date
- 2026-07-17
AI Technical Summary
Existing data lineage analysis methods are prone to problems such as missing field mapping relationships, unexpanded dynamic structured query language, and incomplete flow paths when dealing with dynamic fragments of structured query language statements, user-defined functions of execution code, and incomplete flow paths. This leads to missing or broken rule parsing lineage results, making it difficult to generate stable fused lineage data.
By acquiring SQL statements, target logs, and metadata for identification, session grouping, and field normalization, and combining lineage resolution services, syntax tree parsing, and flow path parsing, rule-based lineage results are generated. AI-enhanced parsing is then used to generate candidate field mapping relationships and candidate lineage dependencies. Metadata, logs, and path verification are then integrated to generate fused lineage data.
It achieves unified processing of missing or broken content in complex structured query language scenarios, generates continuous fused lineage data, supports subsequent semantic vector generation, vector index establishment, similarity calculation and graph construction, and improves the stability and accuracy of data lineage analysis.
Smart Images
Figure CN122412451A_ABST