Identifikatormapping

Identifikatormapping is the process of linking identifiers used by different information systems that refer to the same real-world entity. It enables data integration, search, and analysis across datasets that employ distinct identifier schemes. Mappings can be one-to-one, one-to-many, or many-to-one, depending on the level of granularity and the relationships defined by domain conventions. They are typically stored in cross-reference tables or managed by identifier registries and mapping services.

Methods for creating and maintaining mappings include manual curation, rule-based approaches, automated matching using string similarity or probabilistic methods, and ontology- or schema-alignment techniques. Persistent identifiers, controlled vocabularies, and ontologies are often used to express relationships such as equivalence, broader, or narrower concepts. Data providers publish cross-links, while aggregators and knowledge-graph platforms consolidate them to improve reach and consistency.

Common sources span multiple domains. In biology, mappings frequently connect identifiers from Ensembl, NCBI Entrez Gene, UniProt, and other databases. In scholarly publishing, mappings relate DOIs, PubMed IDs, and Crossref metadata. In government and commerce, identifiers such as ISINs, CIKs, or GTINs may be linked across systems.

Quality and governance are central concerns. Mapping quality depends on accuracy, coverage, update frequency, and traceable provenance. Conflicts can arise when different mappings imply different relationships or when sources have inconsistent evidence. Licensing and privacy considerations may also affect shareability.

Applications include data integration, knowledge-graph construction, data cleaning, and reproducible research, where reliable identifier mappings support interoperability across heterogeneous data sources.