Converting BIBFRAME to Schema.org

This short presentation at the European BIBFRAME Workshop 2018 in Florence, Italy explains how we at the National Library of Finland are converting records of the Finnish National Bibliography Fennica to the Schema.org model for publishing, by first converting them to BIBFRAME and then further to Schema.org.

Google Slides: https://tinyurl.com/bf-to-schema

Transcript

1. Converting BIBFRAME to Schema.org Osma Suominen European BIBFRAME Workshop Florence, Italy 18 September 2018
2. Why? 1. Schema.org has become the lingua franca of the Web 2. Big picture view of our data: entities and their relationships 3. KISS principle It is useful to go via BIBFRAME, but for our current use cases and experiments, Schema.org seems like a better fit
3. MARC 1M records BIBFRAME Raw Schema.org Linked Schema.org 40M triples marc2bibframe2 converter bf-to-sparql.rq SPARQL CONSTRUCT query reconcile.rq SPARQL CONSTRUCT query
4. bf:Work f:000095841#Work240-15 bf:Work, bf:Text f:000095841#Work bf:translationOf “A brief history of time” rdfs:label bf:Language lclang:eng bf:language “A brief history of time suomi” rdfs:label bf:Language _:b0 bf:language bf:Identifier lclang:fin bf:identifiedBy bf:Instance f:000095841#Instance “Ajan lyhyt historia : alkuräjähdyksestä mustiin aukkoihin /” rdfs:label bf:instanceOf BIBFRAME 2.0
5. bf:Work, schema:CreativeWork f:W00009584101 bf:Work, schema:CreativeWork f:W00009584100 schema:translationOfWork “A brief history of time” schema:name schema:inLanguage “Ajan lyhyt historia : alkuräjähdyksestä mustiin aukkoihin” schema:name bf:Instance, schema:CreativeWork f:I00009584100 schema:exampleOfWork “eng” schema:workTranslation “fin” schema:inLanguage schema:workExample “Ajan lyhyt historia : alkuräjähdyksestä mustiin aukkoihin” schema:name Raw Schema.org
6. bf:Work, schema:CreativeWork f:W00009584101 bf:Work, schema:CreativeWork f:W00009584100 schema:translationOfWork “A brief history of time” schema:name schema:inLanguage “Ajan lyhyt historia : alkuräjähdyksestä mustiin aukkoihin” schema:name bf:Instance, schema:CreativeWork f:I00009584100 schema:exampleOfWork “en” schema:workTranslation “fi” schema:inLanguage schema:workExample “Ajan lyhyt historia : alkuräjähdyksestä mustiin aukkoihin” schema:name BCP 47 language tag BCP 47 language tag Reconciled Schema.org
7. bf-to-sparql.rq query (328 lines)
8. Observations and lessons learned 1. The conversion is relatively simple on a logical level, but the devil is in the details 2. The LOC marc2bibframe2 converter makes a good job of extracting information from a MARC record, but: a. many aspects are still just strings after conversion b. “one record at a time” conversion requires separate merging of works c. identity management is a real challenge, as the underlying MARC records may change at any time
9. Thank you! Questions? osma.suominen@helsinki.fi - @OsmaSuominen http://data.nationallibrary.fi - @NatLibFiData Code: https://github.com/NatLibFi/bib-rdf-pipeline https://github.com/NatLibFi/bib-lod-ui These slides: http://tinyurl.com/bf-to-schema