WikiCite 2017 was a 3-day conference, summit and hack
day hosted in Vienna, Austria, on May 23-25, 2017. It expands efforts
started with WikiCite
2016 to design a central bibliographic repository, as well as tools and
strategies to improve information quality and verifiability in Wikimedia
projects. Its goal is to bring together Wikimedia contributors, data modelers,
information and library science experts, software engineers, designers and
academic researchers who have experience working with Wikipedia's citations and
Wikidata and BOOOOKS
Books are complex: you can understand what a book is on many different levels
Simplify FRBR to just two levels ("work" and "edition"), and ignore the item and manifestation levels of FRBR.
Other ontologies (BibFrame) and library projects are also converging to a 2-level system (e.g. OpenLibrary, World Cat).
Books and Wikidata: initially, Wikidata items about books were created because we had Wikipedia articles about them ("work" level). Then, we have Commons, Wikisource, Wikiquote and the like, which are at the "edition" level. Citations are also at the edition level (specific page, ISBN etc.) If you have a page number, it's an edition, not a work.
Problem: how should we model the relationship between "work" and "edition" levels on Wikidata? Examples: 2 items for 1 book = consistent but complicated; 1 item per book = simple, maybe too much; mixed approach = maybe right, surely messy.
- Important/notable books: are likely to have many editions and translations, and confusion may ensue.
- For less notable book: If there's only 1 Edition, should we still make an item for Work?
- How about Anthologies (collective editions including several works)?
How to model derivative works (e.g. movies based on books)?
Same property can be used for different scopes (e.g. 'language' for the edition, vs the original language of the work).
Question from the audience: what's the solution Wikidata community came with for DOI's for books? Answer: there's no solution. We need to discuss it.