Wikidata:WikidataCon 2017/Notes/FRBR and Ontologies of Popular Music in Wikidata
Title: FRBR and Ontologies of Popular Music in Wikidata
Note-taker(s): Thiemo
Speaker(s)[edit]
Name or username: Robert Fernandez, WMDC Contact (email, Twitter, etc.): @wikigamaliel wikigamaliel@gmail.com
Useful links: https://commons.wikimedia.org/wiki/File:FRBR_and_ontologies_of_popular_music_in_Wikidata.pdf
Abstract[edit]
This presentation will discuss FRBR's framework for organizing popular music and its limitations as a potential model for the organization of popular music in Wikidata. It will discuss how popular music is currently organized in Wikidata and how that organizational structure is inadequate to deal with both current and future needs. Since Wikidata is still flexible enough to accommodate structural changes to its ontologies, these issues can be addressed now before such changes become more difficult or large data sets regarding popular music are imported and it is found that existing ontologies are inadequate to fully leverage this data.
Holden, Christopher. "The Definition of the Work Entity for Pieces of Recorded Sound." Cataloging & Classification Quarterly, vol. 53, no. 8, Nov-Dec2015, p. 873. doi:10.1080/01639374.2015.1057886; The Definition of the Work Entity for Pieces of Recorded Sound (Q34147945)
Kishimoto, Kevin and Tracey Snyder. "Popular Music in FRBR and RDA: Toward User-Friendly and Cataloger-Friendly Identification of Works." Cataloging & Classification Quarterly, vol. 54, no. 1, Jan. 2016, pp. 60-86. doi:10.1080/01639374.2015.1105898 ; Popular Music in FRBR and RDA: Toward User-Friendly and Cataloger-Friendly Identification of Works (Q34147953)
Collaborative notes of the session[edit]
If you don't know what FRBR is, don't worry.
How many librarians are here? (Laughter.)
All this started at WikiCite.
This is about records being recorded in Wikidata. We still have time to do this right before it's done.
This is about popular music. Comparison to classical music is made, as a counter-example.
Popular music is like an afterthought.
FRBR = Functional Requirements for Bibliographic Records.
Elements (W-E-M-I):
Work
Expression (realizations)
Manifestation (physical embodiment)
Item (unique physical object, a singular exemplar)
Books are brougth up for comparison. This hierarchy is also applied to classical music, via examples. Now applied on popular music: Work: "Dark Side of the Moon". Expressions: The 1972–73 album, or a concert performance. Manifestation: 1973 LP release, an other one with a green record, and so on.
W-E-M-I applied on Wikidata:
How to model all this, especially the relations?
How to make users understand the differences?
Wikidata currently doesn't model most of the required relations, e.g. from specific expressions to the work.
Covers are mashed together.
P144 (based on) exists. Is this sufficient?
What properties do we need on the manifestation level?
Questions / Answers[edit]
Remark: Wasn't FRBR created to solve exactly what you want to solve? Remark: Some Librarieans start to reject FRBR because sometimes it does not work.
Q: Who is the creator? The composer? Good question.
Q: FRBR is failed to work. I agree. It's used as a starting point for discussion.
Q: Is this about citation? Or to model relationships? That would be very different. What is the use case? Can't be only citation.
Remark: The complexity of the model is there to be able to express everything, but still be discoverable.
Remark: Look into Musicbrainz. They have working solutions already.