IMLS Digital Collections Registry: Description of the XML Export Formats

For people in a hurry to dive right into the data: METS-based Format Custom Format

An experimental XML-based export format is being developed in support of the IMLS Digital Collections Registry. A human-readable HTML browse interface into the registry is also available. This XML format is based on our IMLS DCC Collection Description Metadata Schema which in turn is based on the UKOLN RSLP Collection Description Metadata Schema and the Dublin Core Collection Description Application Profile. In addition, some elements from the vCard standard have been used, plus some custom elements needed specifically for this registry.

Because our metadata schema is composed of several different types of entities and their realtionships to each other, METS seemed like a natural choice for bundling the descriptive metadata for each of the entities into a single file and expressing the structural relationship between these entities. In addition to the METS bundle, we also developed our custom XML schema for bundling these multiple entities into a single file. However, METS provides a much richer framework for bundling and expressing structure than our custom approach. For completeness and comparison both will be presented.

Several XML Schemas were developed in support of this profile. Both the METS and our custom approach utilize these XML Schema. NOTE: These XML Schema are still in development, and several provisional namespaces have been used until more permanent namespaces can be assigned. Several of these XML Schema are based on the XML Schemas for Qualified Dublin Core. However, instead of linking to or importing these schema, local copies have been created. This is because some of the canonical schema hosted at the Dublin Core site are somewhat out-of-date, plus in some cases minor changes were required to support our data. Once these canonical schema have been updated to include all of the newer DC elements required by the profile, and our own data has been cleaned to conform to these schema, we anticipate linking to them instead of hosting our own copies. Without further ado, here are the XML Schemas which are mounted locally:

Namespace	Schema	Brief Description
http://imlsdcc.grainger.uiuc.edu/profile#	IMLSDCCProfile.xsd	This is the root XML Schema for the main collection description profile. It declares several top-level container elements, plus a few new terms needed specifically for the IMLSDCC Profile. It imports from the following other schema:
http://imlsdcc.grainger.uiuc.edu/types#	IMLSDCCTypes.xsd	This XML Schema declares several custom types (mostly enumerated values) that are used by the IMLSDCC Profile. This schema includes several other schema which are dynamically derived from the controlled vocabulary tables in our database: Temporal Coverage, Spatial Coverage, Collection Type, Metadata Schema, Interactivity, Relation Supplement, and Audience.
http://www.w3.org/2001/vcard-rdf/3.0#	vCard.xsd	This XML Schema declares terms from the vCard standard, based on the vCard RDF Schema hosted by the W3C . vCard terms are used for describing people and insitutions.
http://purl.org/dc/elements/1.1/	DC.xsd	This XML Schema declares the basic 15 terms of Dublin Core. It is based on the canonical XML Schemas for Qualified Dublin Core.
http://purl.org/dc/terms/	DCTerms.xsd	This XML Schema declares the qualified Dublin Core terms and types. It is also based on the canonical XML Schemas for Qualified Dublin Core.
http://purl.org/dc/dcmitype/	DCMIType.xsd	This XML Schema declares values used to describe the type of a resource. It is also based on the canonical XML Schemas for Qualified Dublin Core.
http://www.loc.gov/marc.relators/	MARCRel.xsd	This XML Schema declares MARC Relator Terms. It is based on the MARC Relator Terms RDF Schema maintained by the Library of Congress. Even though the XML Schema contains all terms, only the 'own' term is used by the Dublin Core Collection Description Application Profile to identify the owner of a collection.
http://example.org/gen/terms#	Gen.xsd	This XML Schema declares several general purpose terms and types which are used by the Dublin Core Collection Description Application Profile, but are not found in any other "official" namespace. This namespace is provisional, and will be changed once a more suitable value is agreed upon.
http://example.org/cld/terms#	CLD.xsd	This XML Schema declares several terms and types needed specifically for collection description which are used by the Dublin Core Collection Description Application Profile, but are not found in any other "official" namespace. This namespace is provisional, and will be changed once a more suitable value is agreed upon.
http://example.org/cld/type#	CLDType.xsd	This XML Schema declares values used to describe the type of a collection. These values are defined in the Dublin Core Collection Description Application Profile, but are not found in any other "official" namespace. This namespace is provisional, and will be changed once a more suitable value is agreed upon.

Namespace

Schema

Brief Description

http://imlsdcc.grainger.uiuc.edu/profile#

IMLSDCCProfile.xsd

This is the root XML Schema for the main collection description profile. It declares several top-level container elements, plus a few new terms needed specifically for the IMLSDCC Profile. It imports from the following other schema:

http://imlsdcc.grainger.uiuc.edu/types#

IMLSDCCTypes.xsd

This XML Schema declares several custom types (mostly enumerated values) that are used by the IMLSDCC Profile. This schema includes several other schema which are dynamically derived from the controlled vocabulary tables in our database: Temporal Coverage, Spatial Coverage, Collection Type, Metadata Schema, Interactivity, Relation Supplement, and Audience.

http://www.w3.org/2001/vcard-rdf/3.0#

vCard.xsd

This XML Schema declares terms from the vCard standard, based on the vCard RDF Schema hosted by the W3C . vCard terms are used for describing people and insitutions.

http://purl.org/dc/elements/1.1/

DC.xsd

This XML Schema declares the basic 15 terms of Dublin Core. It is based on the canonical XML Schemas for Qualified Dublin Core.

http://purl.org/dc/terms/

DCTerms.xsd

This XML Schema declares the qualified Dublin Core terms and types. It is also based on the canonical XML Schemas for Qualified Dublin Core.

http://purl.org/dc/dcmitype/

DCMIType.xsd

This XML Schema declares values used to describe the type of a resource. It is also based on the canonical XML Schemas for Qualified Dublin Core.

http://www.loc.gov/marc.relators/

MARCRel.xsd

This XML Schema declares MARC Relator Terms. It is based on the MARC Relator Terms RDF Schema maintained by the Library of Congress. Even though the XML Schema contains all terms, only the 'own' term is used by the Dublin Core Collection Description Application Profile to identify the owner of a collection.

http://example.org/gen/terms#

Gen.xsd

This XML Schema declares several general purpose terms and types which are used by the Dublin Core Collection Description Application Profile, but are not found in any other "official" namespace. This namespace is provisional, and will be changed once a more suitable value is agreed upon.

http://example.org/cld/terms#

CLD.xsd

This XML Schema declares several terms and types needed specifically for collection description which are used by the Dublin Core Collection Description Application Profile, but are not found in any other "official" namespace. This namespace is provisional, and will be changed once a more suitable value is agreed upon.

http://example.org/cld/type#

CLDType.xsd

This XML Schema declares values used to describe the type of a collection. These values are defined in the Dublin Core Collection Description Application Profile, but are not found in any other "official" namespace. This namespace is provisional, and will be changed once a more suitable value is agreed upon.

There are also several external XML Schemas which are used or imported by the other schema:

Namespace	Schema	Brief Description
http://www.loc.gov/METS/	METS.xsd	This is version 1.4 of the METS XML Schema.
http://www.w3.org/1999/xlink	XLink.xsd	This schema declares the various attributes used by the XLink standard. These attributes are used by the METS and vCard XML Schema.
http://www.w3.org/XML/1998/namespace	XML.xsd	This schema declares the various special attributes used by XML, such as xml:lang, xml:space,and xml:base.

Namespace

Schema

Brief Description

http://www.loc.gov/METS/

METS.xsd

This is version 1.4 of the METS XML Schema.

http://www.w3.org/1999/xlink

XLink.xsd

This schema declares the various attributes used by the XLink standard. These attributes are used by the METS and vCard XML Schema.

http://www.w3.org/XML/1998/namespace

XML.xsd

This schema declares the various special attributes used by XML, such as xml:lang, xml:space,and xml:base.

Now for some links to the actual data:

First there is the top-level METS-based record: METS.asp. This link will return a list of XLinks to all the NLG Collection Descriptions in the database.

If you follow one of these links, such as METS.asp?2409, you will get to the detailed METS record for the collection, including all of the related entities and their relationships. Detailed description for the collection itself and for all related entities, such as People, Projects, and Institutions, are included in the METS record in seperate METS file elements. Detailed descriptions for related collections are not included in the METS file, but XLink references to their own detailed METS records are included in the METS structMap which defines the relationships between all the entities which make up the complete collection description.

We have also developed a custom record format which can be used in place of METS: Profile.asp. Just Like the METS example, this link will return a list of all the NLG Collection Descriptions in the database.

Also like the METS example, if you follow one of these links you will get the detailed record for a particular collection, such as Profile.asp?2409. The custom format includes detailed descriptions for the collection and all related entities, except other collections. However, there is nothing equivalent to the METS structMap for the custom format. Instead relationships between entities is implicit in the detailed descriptions themselves, which have various relation elements which tie together the different entities, such as the manager of a collection is expressed as:

  <imlsdcc:managedBy xsi:type="dcterms:URI">
    http://imlsdcc.grainger.uiuc.edu/Registry/Person/?761
  </imlsdcc:managedBy>

and in the same file that person is identified as:

  <vcard:UID vcard:TYPE="URI">
    http://imlsdcc.grainger.uiuc.edu/Registry/Person/?761
  </vcard:UID>

or the project which created the collection is expressed as:

  <dc:creator xsi:type="dcterms:URI">
    http://imlsdcc.grainger.uiuc.edu/Registry/Project/?2690
  </dc:creator>

and later in the file that project is identified as:

  <dc:identifier xsi:type="dcterms:URI">
    http://imlsdcc.grainger.uiuc.edu/Registry/Project/?2690
  </dc:identifier>

You can also see that every entity has a unique URI identifier, and it is possible to get a detailed description of just a specific entity without the related entities by dereferencing that URI, such as:

The IMLS Collection Registry is a product of IMLS Digital Collections and Content, a National Leadership Grant funded collaboration among the University of Illinois Library, the Graduate School of Library and Information Science, and the Institute of Museum and Library Services, a Federal agency that fosters innovation, leadership, and a lifetime of learning.