Background Information on the BIBFRAME Ontology Structure

Last updated
Save as PDF
Share
1. Share
2. Tweet
3. Share

The Library of Congress developed the BIBFRAME (Bibliographic Framework) ontology, a model for expressing and sharing bibliographic data in a Linked Open Data environment. It replaces the MARC format with a more web-friendly, semantic structure using RDF (Resource Description Framework), which enables data to be expressed in triples: subject-predicate-object.

In the context of Linked Data and the Semantic Web, an ontology defines a structured framework for organizing information. It includes:

Classes (types of things)
Properties (relationships and attributes)
Rules about how these elements relate to each other

Official Ontology Resources

To learn the ontology and the structure as developed by the Library of Congress, see:

BIBFRAME Ontology (RDF/HTML): https://id.loc.gov/ontologies/bibframe.html
BIBFRAME LC Extension (BFLC): https://id.loc.gov/ontologies/bflc.html

Learn More

For more detailed information, you can explore the following resources at the Library of Congress:

BIBFRAME Overview and Vocabulary — The Library of Congress
BIBFRAME FAQs — Helpful explanations and comparisons to MARC
BIBFRAME cataloging workshop — The Library of Congress

Core Concepts in BIBFRAME

Class

A class defines a group of resources with common characteristics. Key BIBFRAME classes include:

bf:Work — A resource reflecting a conceptual essence (e.g., a novel).
bf:Instance — A material embodiment of a Work (e.g., a specific edition).
bf:Item — A single, physical or digital copy of an Instance.
bf:Agent — A person, organization, or meeting associated with a resource.
bf:Title — A title associated with a Work or Instance.

Classes are identified in the ontology by their type and by the fact that the name of the class begins with an uppercase letter.

Subclass

A subclass is a more specific type of a class, inheriting properties from another class.

All instances of the subclass are also instances of the superclass, and the relations are inherited (Superclass = the "parent" of all the subclasses).

Examples:

bf:VariantTitle is a subclass of bf:Title, used for variants to the Title of the resource.
bf:TransliteratedTitle is a subclass of bf:VariantTitle, used to describe a more specific type of a variant title.

Property

A property defines a relationship between resources or between a resource and a literal value.

There are two main types of properties in use within the BIBFRAME ontology:

Object Property: Links a resource to another resource (attribute).
Datatype Property: Links a resource to a literal value.

In BIBFRAME, properties always serve as predicates in RDF triples, defining relationships or attributes between resources. This aligns with RDF's core model of subject-predicate-object.

Properties are identified in the ontology by their type and by the fact that the name of the class begins with a lowercase letter.

Examples:

bf:title – Connects a Work or an Instance to a bf: Title.
bf:contribution – Links a Work to a bf: Agent who contributed to it.
bf:identifiedBy – Associates a resource with an identifier (e.g., ISBN).

Sub-Property

A sub-property is a more specific version of another property. If a different relation is not defined, this type of relation also indicates inheritance between properties.

Examples:

bf:accompanies is a sub-property of bf:relatedTo
bf:replacedBy is a sub-property of bf:succeededBy (that is a sub-property of bf:relatedTo)

Used with (Domain) and Expected Value (Range)

Used With (domain): The class a property applies to.

Expected Value (range): The expected type of value (another class or a literal).

Used with and Expected Value are defined for properties.

Datatype properties always have an expected value of “rdf-schema#Literal” as they can only receive a text input by definition.
The "used with" and "expected value" are not mandatory. If the relation is blank or defined as “rdf-schema#Resource”, then there is no constraint on the usage of the property with different classes.
The ontology uses the terms "domain" and "range." Alma uses the terms "used with" and "expected value" as these are more user-friendly and explain the relations between the entities more intuitively.

Example:

The property bf:descriptionLevel is used with the class bf:AdminMetadata and has the class bf:DescriptionLevel as the expected value