Skip to main content

What Is the Data Model?

The Data Model defines every metadata field that exists in your archive — what shows up on documents, photographs, artefacts, audio items, video items, and at the collection level. It's seeded with archival-standards defaults (ISAD(G), DACS, Dublin Core, Spectrum) and fully customisable.

Find it at Settings → Data Model.

Data Model

Why it matters​

  • The wizard surfaces these fields — when you create an accession, the Custom Fields step asks about each field your model has defined.
  • The item review screen renders these fields — what you see when you expand a row on the Items tab.
  • The AI populates these fields — what the model extracts is shaped by which fields are active and what they expect.
  • Exports map from these fields — every format's mapping starts here.

Change the Data Model and every part of the platform that touches metadata updates accordingly.

How it's organised​

The Data Model is split by category:

  • Shared — fields that appear on every category (Title, Date, Creator, Description, Subjects, …)
  • Document — bibliographic fields (publisher, language, page count, …)
  • Photograph — image-specific fields (caption, photographer, technique, …)
  • Artefact — conservation-style fields (materials, dimensions, condition, …)
  • Audio — audio-specific fields (runtime, format, sample rate, transcript, …)
  • Video — video-specific fields (runtime, resolution, codec, transcript, …)
  • Collection — fond / sub-fond / series level fields (scope and content, biographical history, …)

Within each category, fields are grouped:

  • Descriptive — what the item is
  • Administrative — access, rights, provenance, sensitivity
  • Conservation — condition, materials, environment (object-focused categories)
  • Technical — format, dimensions, technical metadata
  • Custom — your organisation's additions

The default schema​

The defaults are derived from archival standards. A few examples:

Documents — Descriptive:

  • Title (required, item / collection scope, Text)
  • Date (recommended, item scope, Date)
  • Creator (recommended, item scope, Text)
  • Description (required, item / collection / series scope, Long text)
  • Subjects (recommended, item scope, List, bound to LCSH)
  • People mentioned (optional, item scope, List, bound to FAST)
  • Places mentioned (optional, item scope, List, bound to TGN)
  • Organisations mentioned (optional, item scope, List, bound to FAST)

Documents — Administrative:

  • Access rights (required, item / collection scope, Text)
  • Sensitivity level (required, item scope, Enum: Public / Restricted / Closed)

Artefacts — Conservation:

  • Condition (recommended, item scope, Long text)
  • Materials (recommended, item scope, List, bound to AAT)
  • Dimensions (optional, item scope, Text)

You can see and edit every default field. Deprecated defaults are hidden by default; toggle Show deprecated to see them.

What you can change​

For every field:

  • Label and help text (cosmetic)
  • Required / Recommended / Optional
  • Vocab binding — switch the bound vocabulary, or unbind to free-text
  • Type — though changing a populated field's type is risky and warns you first

For every category, you can:

  • Add new fields in any group
  • Deprecate unused fields (hidden but data preserved)
  • Apply default vocab bindings to restore the seed bindings if you've drifted
  • Reset to defaults for the whole category (offers a backup first)

Org-level vs personal​

On Community and Professional, Data Model changes are personal — they affect your account only.

On Team and Enterprise, Data Model is org-wide. Only Admins can edit it. Everyone in the org sees the same schema.

This is intentional: a shared Data Model is critical for consistency in a multi-seat archive.

Migrating an existing Data Model​

If you're moving from another catalogue and have an existing schema definition (CSV, JSON, ArchivesSpace template), you can import it via Settings → Exports → Import Profiles → Data Model template. The import shows a diff against your current model — you decide what to merge.

What's next?​