DCMIThe DCMI Abstract Model (DCAM) is a foundational specification from the Dublin Core Metadata Initiative that defines the information structures underlying all Dublin Core metadata. Published as a DCMI Recommendation in June 2007, it provides a syntax-independent framework for understanding what Dublin Core metadata descriptions actually contain and how their components relate to one another. The model serves as a bridge between the conceptual intent of Dublin Core and its various concrete serializations in XML, RDF/XML, HTML, and other formats.
Background
As Dublin Core metadata grew from a simple fifteen-element vocabulary in the mid-1990s into a richer ecosystem of properties, classes, and encoding schemes, the need for a formal information model became increasingly clear. Different encoding syntaxes -- HTML meta tags, XML, RDF/XML -- were expressing Dublin Core metadata in different ways, and without a shared abstract model, interoperability between these encodings was difficult to reason about.
The DCMI Abstract Model was developed by Andy Powell, Mikael Nilsson, Ambjorn Naeve, Pete Johnston, and Tom Baker, drawing heavily on the conceptual underpinnings of the W3C Resource Description Framework (RDF). It was issued as a DCMI Recommendation on 4 June 2007, with minor errata corrections in September 2007 and February 2013. The model formalized concepts that the Dublin Core community had been developing informally through earlier documents such as the DCMI Grammatical Principles.
Purpose & Scope
The DCMI Abstract Model serves three primary audiences: developers building software that processes Dublin Core metadata, authors of new encoding guidelines for Dublin Core, and designers of metadata application profiles based on DCMI vocabularies. Its purpose is to define the components and constructs used in Dublin Core metadata at a level above any particular serialization format, enabling better mappings and cross-syntax translations.
The model does not define a formal semantics of its own. Instead, it defers to RDF and RDF Schema semantics, providing a table of equivalences between DCAM concepts and their RDF counterparts.
Key Elements / Properties
The DCMI Abstract Model defines three interrelated sub-models:
The Resource Model
Describes the fundamental relationship between resources and their descriptions:
| Construct | Definition |
|---|---|
| Resource | Anything that might be identified |
| Property-value pair | A property and a value used to describe a resource |
| Literal value | A value which is a literal (Unicode string with optional language tag or datatype) |
| Non-literal value | A value which is a physical, digital, or conceptual entity |
The Description Set Model
Defines how metadata statements are structured:
| Construct | Definition |
|---|---|
| Description set | A set of one or more descriptions |
| Description | One or more statements about a single resource |
| Statement | An instantiation of a property-value pair |
| Value surrogate | A literal or non-literal value surrogate representing the value |
| Value string | A literal that encodes or represents a value |
The Vocabulary Model
Defines the types of terms used in Dublin Core vocabularies:
| Construct | Definition |
|---|---|
| Vocabulary | A set of one or more terms |
| Term | A property, class, vocabulary encoding scheme, or syntax encoding scheme |
| Property | A specific aspect or attribute used to describe resources |
| Class | A group of resources with shared characteristics |
| Vocabulary encoding scheme | An enumerated set of resources |
| Syntax encoding scheme | A set of strings with mapping rules to a set of resources |
Serializations & Technical Formats
The DCMI Abstract Model is deliberately encoding-independent. It does not prescribe a particular serialization but instead defines the abstract structures that concrete encodings must represent. Separate DCMI encoding guidelines specify how the model maps to specific formats including HTML meta tags, XML, and RDF/XML.
The DCAM namespace URI is http://purl.org/dc/dcam/, which defines terms such as memberOf and VocabularyEncodingScheme that are used in RDF representations of Dublin Core metadata.
Governance & Maintenance
The DCMI Abstract Model is maintained by the Dublin Core Metadata Initiative, an organization supporting innovation in metadata design. The specification was developed with input from the DC Usage Board and the DCMI Architecture Community. Changes follow the DCMI recommendation track process. The document is licensed under a Creative Commons Attribution 4.0 International License.
The model has remained stable since its 2007 publication, reflecting its role as a conceptual foundation rather than a frequently revised vocabulary.
Notable Implementations
The DCMI Abstract Model underpins all implementations of Dublin Core metadata. It is implicitly relied upon by any system that uses DCMI Metadata Terms, whether in library catalogs, institutional repositories, digital asset management systems, or open government data portals. The model's description set concept directly influenced the design of the Description Set Profile constraint language used in Dublin Core Application Profiles.
Systems that perform cross-walk mappings between Dublin Core and other metadata standards (MARC, MODS, EAD) depend on the abstract model's definitions to ensure semantic fidelity across encodings.
Related Standards
- Dublin Core Metadata Element Set (DCMES) -- the original fifteen-element vocabulary that the abstract model formalizes
- DCMI Metadata Terms -- the full set of properties, classes, and encoding schemes defined by DCMI, all grounded in this abstract model
- RDF / RDF Schema -- the W3C standards on which the DCAM's formal semantics are based
- DCMI Grammatical Principles -- the predecessor document that introduced many of the concepts later formalized in the abstract model