METS

METS (Metadata Encoding and Transmission Standard) is an XML schema, maintained by the Library of Congress, for packaging together the descriptive, administrative, and structural metadata that describes a digital object, along with (or with links to) the object’s own content files. It functions as a standardised “wrapper” format rather than a content-description vocabulary in its own right.

Scope note

METS is fundamentally a digital-library and digital-preservation systems specification, and most in-depth METS documentation (schema internals, profile registration, implementer guidance) is aimed squarely at library-systems developers rather than research administrators. The angle kept here is deliberately narrow: why a repository manager or CRIS professional is likely to meet METS at all, and what it is doing when they do.

The section structure, and why a repository manager meets it

A METS document organises a digital object into named sections: dmdSec (descriptive metadata — bibliographic information about the object), amdSec (administrative metadata, itself subdivided into technical, rights, source/provenance, and digital-provenance subsections), fileSec (a manifest of the actual content files), structMap (the logical or physical structure tying files and metadata sections together — e.g. page order in a digitised book), plus optional structLink and behaviorSec sections. Rather than replacing established metadata standards, METS is deliberately designed to embed or reference them within its sections — MODS for descriptive metadata, PREMIS for preservation metadata, MIX for image-technical metadata, textMD for text-technical metadata — functioning as connective packaging around standards that already exist, not a competing description vocabulary. A research-data or institutional repository that ingests digitised archival material, or that exports an Archival Information Package for long-term preservation, is likely to encounter METS at exactly that boundary — as the packaging format a digital-preservation system like Archivematica or RODA produces or consumes.

Worked example / counter-example

A digitisation workflow at a national library wraps a scanned manuscript’s page images, OCR text files, MODS descriptive metadata, and PREMIS preservation-event metadata inside a single METS document, with a structMap defining the correct page sequence — that packaged object is what moves through an Archival Information Package pipeline as one unit. A bare PDF with no accompanying metadata sections is not METS, and a standalone Dublin Core metadata record, with no wrapper tying it to specific content files or structure, is not METS either — METS would express that same descriptive record inside a dmdSec, alongside the file and structural information it is currently missing.

References

• Library of Congress, “METS Schema and Documentation” (loc.gov/standards/mets).