Glossary · System coordination, integration and orchestration
Schema mapping
Also known as: Data mapping
German: Schemaabbildung
In data integration, schema mapping is the specification of how the elements of one data schema correspond to the elements of another, including transformations of names, structures, data types and units, so that data can be exchanged or merged between systems.
- System integration
In one sentence
Schema mapping specifies how elements of one data schema correspond to another, including conversions of names, types and units.
Example
A mapping translates the ERP field MATNR to the MES attribute materialId and converts the target weight from kilograms to grams for the filling machine.
How it applies
- Engineering: A mapping covers structure (which field goes where), semantics (does "status" mean the same thing on both sides?) and values (units, code lists, time zones). Semantic mismatches are the hardest to spot.
- Maintenance: Mappings break when either schema changes. Version them together with the schemas and include them in Schema governance.
- Verification: Test mappings with real data, edge cases and missing values; silent truncation or unit errors can go unnoticed for a long time.
- Documentation: Keep mappings in a readable, reviewable form (mapping tables, not only code). In technical documentation, schema mappings connect content metadata models, for example a company's product taxonomy to iiRDS metadata.
Schema mapping vs. protocol conversion
Protocol conversion changes the transport of data between communication protocols. Schema mapping changes the structure and meaning of data between models. Integration projects usually need both, and mixing them in one undocumented step makes errors hard to trace.