X-Data Standard uses a large model and semantic analysis to identify core fields in selected data assets, then automatically extracts lookup table definitions, generates data standard definitions, and recommends standard mappings.
Prerequisites
Before you begin, ensure that you have:
Purchased both the Data Standard and the X-Data Standard function plans.
Configured and enabled X-Data Standard. For more information, see .
Permissions
Super administrators, Data Standard administrators, and custom global roles with the Standard-Manage permission can use X-Data Standard.
Flow for extracting lookup table definitions
The large model extracts lookup table definitions through four stages: Configure data scope > Identify core fields > Explore and sample data > Extract lookup table definitions.
Data scope: Select the data scope from which to extract lookup table definitions.
Identify core fields: The large model analyzes the semantics of the selected data assets to identify core fields for the next step.
Explore and sample data: The system samples and explores the identified core fields to understand their data distribution.
Extract lookup table definitions: Based on data asset metadata, exploration results, and sample data, the large model generates lookup table definitions. Each definition includes a name, code, and description for the lookup table, along with code values, code names, and optional code descriptions.
Standard definition flow for AI recommendation
The large model extracts data standard definitions through four stages: Configure data scope > Identify core fields > Explore and sample data > Extract data standard definitions.
Data scope: Select the data scope from which to extract data standard definitions.
Identify core fields: The large model analyzes the semantics of the selected data assets to identify core fields for recommendations.
Explore and sample data: The system samples and explores the identified core fields to understand their data distribution.
Extract data standard definitions: Based on asset metadata, exploration results, and sample data, the large model generates data standard definitions. Each definition includes business properties (data standard code, name, English name, type), technical properties (data type, is unique value, is nullable, range), the data standard set and folder, effective period, owner, description, mapping monitoring configuration, intelligent mapping configuration, and associated information.
Flow for recommending standard mappings
The large model recommends standard mappings through three stages: Configure data scope > Identify core fields > Recommend standard mappings.
Data scope: Select the data scope for which to recommend standard mappings.
Identify core fields: The large model analyzes the semantics of the selected data assets to identify core fields for recommendations.
Recommend standard mappings: The system parses metadata, exploration results, and sample data, then matches fields to data standards using the existing data standard definitions.