BIBFRAME AV Assessment: An Introduction

Why an AV Assessment Is Needed

AV resourcesfilm, video, audio recordings, streaming media, and newer digital formatshave distinct metadata requirements. Traditional cataloging standards often struggled to capture:

• Technical specifications (codec, resolution, bitrate)
• Production roles (director, producer, cinematographer)
• Rights and licensing information
• Versioning and format migration histories
• Usergenerated metadata such as subtitles or transcripts

Because BIBFRAME is expressed as linked data, it offers the potential to model these complexities more flexibly, but only if the model is rigorously tested against realworld AV collections. The assessment therefore serves three core purposes:

• Identify gaps between the model and current cataloging practice.
• Propose extensions or refinements to the BIBFRAME ontology.
• Provide guidance for libraries that wish to adopt BIBFRAME for AV collections.

Key Components of the Assessment

1. Data Mapping Exercises

Catalogers select a representative sample of AV items and map existing MARC 21 fields to BIBFRAME entities and properties. The exercise highlights fields that translate cleanly (e.g., 245 title, 260/264 publication info) and those that do not (e.g., 007 fixed fields for technical characteristics).

2. Ontology Gap Analysis

Using tools such as SHACL (Shapes Constraint Language), evaluators test whether the BIBFRAME schema can express required AV concepts without creating adhoc workarounds. Missing classes (e.g., bf:Codec) or properties (e.g., bf:runtime) are documented.

3. Interoperability Testing

Linkeddata exchange scenarios are simulated between libraries, archives, and streaming platforms. These tests verify that BIBFRAME records can be harvested, enriched, and displayed in external environments while preserving essential AV details.

4. UserCentric Evaluation

Enduser taskssuch as searching for a movie by director or filtering videos by resolutionare performed on prototype discovery interfaces that consume BIBFRAME AV data. Feedback informs whether the model supports realistic discovery workflows.

Findings to Date

Several pilot projects across North America and Europe have contributed to a growing body of evidence. The most common observations include:

• Technical Metadata Gaps: BIBFRAME lacks dedicated properties for many codec and container attributes. Extensions like bf:AudioCodec and bf:VideoCodec have been proposed.
• Complex Creatorship: AV works often involve multiple, hierarchical creative roles (e.g., executive producer vs. associate producer). The current bf:contribution hierarchy does not capture this granularity without custom qualifiers.
• Rights Representation: While BIBFRAME can link to rights statements, it does not inherently model timebound licenses or platformspecific restrictions, a critical need for streaming services.
• Version Control: The model distinguishes between bf:Instance and bf:Work but lacks a straightforward way to express version histories such as remastered edition or directors cut.
• Discovery Flexibility: When AV metadata is exposed as linked data, faceted search (by format, runtime, language) becomes more powerful, confirming one of the major benefits of the BIBFRAME approach.

Recommendations for Enhancing BIBFRAME AV Support

1. Introduce a Dedicated AV Module: A separate namespace (e.g., bfav:) can house classes like bfav:VideoInstance, bfav:AudioInstance, and properties for technical specifications.
2. Standardize Role Taxonomy: Adopt an extensible role vocabulary (such as the FRBRR or the IPRC roles) and map it to BIBFRAME contribution properties.
3. Model Rights as FirstClass Entities: Use linkeddata patterns to connect a work or instance to rights statements that include temporal and jurisdictional qualifiers.
4. Versioning Framework: Define relationships like bfav:hasVersion and bfav:isVersionOf to express edition evolution and format migration.
5. Implement Validation Shapes: Publish SHACL shapes that libraries can use to validate AV records against the enhanced schema, ensuring data quality.
6. Provide Migration Guides: Offer stepbystep documentation for converting MARC 21 AV fields (e.g., 007, 300, 511) into BIBFRAME equivalents.

Next Steps for Libraries

Libraries interested in adopting BIBFRAME for AV collections can begin with a phased approach:

1. Pilot a Small Collection: Choose a subset of items (e.g., a local film series) and create BIBFRAME records using existing tools like BIBFRAME Editor.
2. Validate Against Assessment Findings: Run SHACL validation to check for missing technical or rights data.
3. Collaborate with Community Projects: Join initiatives such as the BIBFRAME AV Working Group or the Linked Data for Libraries (LD4L) consortium to share solutions.
4. Iterate and Extend: Incorporate the recommended AV module, test with discovery interfaces, and refine based on user feedback.

By treating the assessment as a living document, libraries can contribute to a more robust, futureproof bibliographic ecosystem that fully embraces the richness of audiovisual material.

Further Reading

• Library of Congress, BIBFRAME: An Introduction, 2022.
• National Information Standards Organization (NISO), Best Practices for AV Metadata, 2023.
• LD4L Working Group, Linked Data for Audiovisual Collections, 2024.
• W3C, SHACL Specification, 2021 (for validation guidance).