The process of creating documented correspondences (mappings or crosswalks) between elements, fields or constructs in different metadata schemas or models to enable data exchange, conversion, aggregation and semantic interoperability.

Principle

Mappings translate or align elements between schemas but vary in expressiveness: they can be direct (one‑to‑one), aggregative (one‑to‑many), decompositional (many‑to‑one) or lossy; therefore a mapping’s adequacy depends on semantic alignment, intended use and documented transformation rules.

Demonstration

Illustrative scenario: A mapping specifies how a simple schema’s 'creator' field corresponds to a richer bibliographic standard’s '100' or '700' fields and whether to preserve name qualifiers or move them to notes. The mapping documentation records where semantics cannot be preserved and how data should be transformed.

Misapplication

Assuming mappings are lossless and symmetric. The semantic error is treating a crosswalk as a perfect, reversible translation; in reality differences in cardinality, controlled vocabularies, or structural expressiveness can cause information loss or require enrichment.

Consequence

Well‑documented mappings enable interoperable data exchange, bulk conversions and integrated discovery; poorly designed or undocumented mappings produce semantic drift, data loss, inconsistent downstream interpretation and maintenance burdens.

Reversal

Where schemas embody fundamentally different conceptual models or ontologies, direct field‑to‑field mapping may be impossible; alternative strategies (use of intermediary models, semantic enrichment, or preservation of provenance and original fields) are required instead of naive crosswalks.

Boundary

Clearly within: explicit field‑to‑field or construct‑to‑construct correspondence tables and transformation rules intended to convert or align records. Boundary case: automated heuristic mapping that proposes correspondences requiring human review. Clearly outside: translating human language in content fields (that is data transformation rather than schema mapping).

Semantic Tension

Interoperability (Broad, Pragmatic Mappings) ↔ Semantic Fidelity (Precise, Lossless Representation): the desire to maximize exchange and compatibility competes with the need to preserve original semantics and nuance.

Synthesis

Metadata mapping is a practical engineering discipline that negotiates between incompatible models: the mapping’s value is measured less by the illusion of perfect translation than by clear documentation of transformations, remaining omissions and provenance to support correct interpretation.