Adam Retter

adam@evolvedbinary.com
 

The National Archives
2020-05-28

@adamretter

• All are recognised and respected standards!

• Major Use

  • Library of Congress PREMIS3 - Record/Agent Entities

  • DCMI Dublin Core Terms - Record/Agent Description

  • W3C PROV - Provenance (and Versioning)

• Minor Use

  • W3C ODRL - Closure and Gradated Access

  • RDA - Record/Agent Description

  • Europeana Data Model - Record Sequencing

• Others

  • FOAF - Agent Description

  • Schema.org - Secondary Identifiers and Dimens.

• Knowing how Gov/TNA managed its records has value:

  • For TNA - When/Why/How did we do that?

  • For Readers - How did Gov attitude about a record change?

  • In itself... considered a Public Record???

• Historically we haven't supported this!

  • PROCat/ILDB - Keeps only the previous change

  • DRI Catalogue - No history. DRI Archive keeps in near-line

  • SAR - History of access constraints

  • Discovery - Does not tell us the custodial history

• Therefore... Catalogue entries must become Immutable!

• Breaking down the walls between
  Catalogue and Archive! :-)

• Public Record

  • As defined by the Public Records Act 1967

  • The physical/digital object itself that is accessioned into the Archive

• The description of the physical/digital object that goes into the Catalogue

• Changes to either the physical/digital object or its description

• Its Identifier?

• Public Record

  • As defined by the Public Records Act 1967

  • The physical/digital object itself that is accessioned into the Archive

• The description of the physical/digital object that goes into the Catalogue

• Changes to either the physical/digital object or its description

• Its Identifier?

1. Separate the mutable properties from the immutable

2. Make a copy of the existing (mutable) description

3. Change the copy of the description

4. Store additional provenance information

• Concept of a Record

  • Immutable descriptive properties of a record - Enduring

  • Has "Descriptions" and "Manifestations"

• Description of a Record

  • Mutable descriptive properties of a record - Transient

  • Represents the description of a record at a point-in-time

• Manifestation of a Record

  • Technical properties of the physical/digital object

  • One for each physical/digital record (or revision of)

  • Also used for Digitisation & Surrogates!

• Digital File

• Every resource in our RDF Graph must have a URI

  • URI has a base followed by a local identifier

• Base Identifier should use organisation's domain name

  • http://cat.nationalarchives.gov.uk

• Need Local Identifiers for:

  • Record - Concept, Description, Manifestation, and Digital File

  • Agent - Person, Group, Family, Organisation, Corporate Body, Hardware, and Software

  • Location

• CCR - Classic Catalogue References (PROCat/ILDB)

  • Convey Knowledge 👍

    • Ambiguous 🤨

  • Non-Uniform 😞

  • May Change 🤫

  • No facility for Revisions (descriptions) or Manifestations 😢

  • Not Suitable for use in URI 😢

• GCR - Generated Catalogue References (DRI)

  • Convey less knowledge than CCR 😞

  • Uniform 👍

  • Never Change 👍

  • No facility for Revisions (has Manifestations) 🤨

  • Not Suitable (as published) for use in URI 😢

• Learn from CCR and GCR and improve!

  • Convey Only Immutable Knowledge

    • Never Change 👍

  • Improved encoding over CCR

    • Shorter Record Numbers 👍

  • Perfect for use in URI 👍

• Different schemes for human vs. machine use

• Suitable for Linked Data use

• Can complement existing CCR or GCR

  • or... Should new records only have an OCI???

• Comprised of Four Components

• Three Immutable Properties for Conveying Knowledge

• Encodes a Record Number

  • OCI Base25 Alphabet

  • Initialised per Creator Reference and Accession Year pair

• Has a fixed component separator (full-stop)

• Syntax

  • Scheme:

    
    

  • Examples:

• Comprised of Five Components

  • First four are the same as Concept of a Record

  • Fifth Component is a Description Number

    • Initialised per Concept of a Record

• Has the same fixed component separator (full-stop)

• Syntax

  • Scheme:

    
    

  • Examples:

• Comprised of Five Components

  • First four are the same as Concept of a Record

  • Fifth Component is a Manifestation Number

    • Initialised per Concept of a Record

• Has the same fixed component separator (full-stop)

• Syntax

  • Scheme:

    
    

  • Examples:

• Concept of a Record

  • Physical

    
    

  • Digital

    
    

• Description of a Record

  
  
  
  

• Manifestation of a Record

• OCI - Digital File

  • Compressed and valid URI

  • Designed for Machines

• ODRL - Open Digital Rights Language

  • Closure

  • Gradated Access?

• Tools we have produced

  • Kettle Jena Plugins - https://github.com/nationalarchives/kettle-jena-plugins

  • OCI Tools Scala - https://github.com/nationalarchives/oci-tools-scala

  • OCI Tools TypeScript - https://github.com/nationalarchives/oci-tools-ts

• Circulate/Publish Information

  • Omega Catalogue Data Model (PDF)

  • Blogs

    • Archival Catalogue Record Identifiers

    • Archival Identifiers for Digital Files

    • Extreme Identifiers (for use in URI)

• Export data from ILDB into OCDM RDF

• Implement (partial) PROCat Replacement

• Stretch goal: import some DRI Born Digital records

• Refine OCDM for use by other Archives ???

  • Is it an alternative to RiC-O?