WorldCat as BIBFRAME:
450 Million Works
650 Million Instances
250 Billion RDF Triples
MARC is re-interpreted as BIBFRAME entities properties and relationships whilst preserving provenance #bfwe2026

WorldCat has been "remodelled as BIBFRAME" with new ingest pipelines, without disrupting existing MARC workflows. #bfwe2026

Where is WorldCat going? - Transition to a connected knowledge graph.
Work is underway to turn WorldCat into a trusted knowledge graph to support discovery and cataloguing use-cases, including AI services #bfwe2026

WorldCat is "a description of the world's collective knowledge" as held by the library community.
The database is agnostic - it can viewed via different lenses: MARC, Dublin Core and BIBFRAME.
#bfwe2026

"Using BIBFRAME data at scale" - Jeff Mixter & Chris Madden, OCLC.
Contextualising the evolution of WorldCat - curated over 50 years by libraries around the world "built with the library community and continuously shaped by it" - not simply 630 million+ MARC records
#bfwe2026

JCricket is designed to integrate rather than replace - it supports record-sharing and interoperability for data and also entity-sharing across systems (including FOLIO and Alma). It also integrates with other editors and applications (Sinopia, Marva, Wikidata)
Creating a knowledge network #bfwe2026

Entities are shared by all ShareVDE users, significantly reducing the amount of effort required to catalogue new instances of existing works meaning far less duplication of effort across libraries #bfwe2026

Cataloguing becomes synchronous, interactive & bidirectional.
JCricket Editor supports external integrations with existing library systems (eg FOLIO) to enable cataloguers to move data from a non-Linked Data source into the ShareVDE environment at either the BIBFRAME Work or Instance level #bfwe2026

JCricket is developed following several drivers for change: the desire to move towards a shared ecosystem of interconnected data that is supported via APIs and Linked Open Data #bfwe2026

The traditional model workflow involves export/import of MARC records & querying via Z39.50 to 'copy' records across to a siloed system #bfwe2026

"From asynchronous exchange to entity-sharing: new paradigms of the bibliographic workflow" - Tiziano Possemato
& Annalisa Di Sabato, Share Family Initiative. #bfwe2026
JCricket as an Operational Entity Editor

This is crucial in a hybrid environment where 'legacy' metadata exists alongside more recent and fully 'linked' metadata - you catalogue will probably never be perfect! #bfwe2026

The quality of the underlying metadata is crucial - when we bridge the gap between data and display, the data we display is only as good as the data that we have!
There are still eclectic genre form terms in the underlying data as well as issues with missing/a lack of data elements #bfwe2026

Relationships between Instances are more clearly demonstrated to end-users - different editions of the same Work are grouped together.
Equally, relationships between Agents and Subjects are made explicit; users can navigate between them easily. #bfwe2026

Minimalistic Work and Instance types, with more detail and complexity expressed as 'Category' types - this brings overlapping schemes and dimensions together #bfwe2026

Data transformation scripts were needed to map the data into the new model, whilst still enabling export as MARC (to support other libraries in Sweden to use the National Library's data) #bfwe2026

Designing a Linked Data search interface required reconciling the principles of BIBFRAME, RDA and national schemas, and adjusting classes, properties and relationships between terms to support the needs of different user groups. #bfwe2026

3 design principles:
1. No more puzzles about Material Type - no need to look at multiple MARC fields to determine that a Book is a Book!
2. Show data as it is - avoid confusion between terminology
3. Make use of Linked Data - make the relationships between entities clear
#bfwe2026

"From Linked Data Cataloging to Linked Data Search" with Lisa Sjögren & Katja Naischtein of National Library of Sweden #bfwe2026
Getting Linked Data working better in the National Library of Sweden for search and discovery, as well as for cataloguing

Policy development is key to help the community manage this shared open metadata ecosystem with robust permissions - no editing wars! #bfwe2026