If we build it, they will come: authority data for a linked data future

Editor's Notes

1. Jennifer: We began our report with a caveat: the big picture that we envision cannot be supported by current systems. Acknowledging this was liberating: it allowed us to dream.
2. Jennifer: For the next 15 minutes, we invite you to dream with us.
3. Jennifer: We conducted our work under the following premises: Linked data is here to stay Linked data has been embraced by billion dollar corporations, internet search engines, international governments, and national libraries abroad; imagine what we could with that power A move to linked data requires a complete remodeling of both bibliographic and authority data For catalogers, that means we must stop thinking about The Record in the monolithic sense and start thinking about data and how data points link to one another to create different views for discovery
4. Jennifer: Interface designers and cataloging policy makers are locked into a chicken-and-egg conundrum Discovery layers and cataloging interfaces don’t provide us with functionality that utilizes authority data because the bulk of that data isn’t rich or granular enough to power a useful service; on the other hand, Those who make cataloging policy--at local, as well as the national levels--ask “why expend resources creating “extra” data if our systems can’t make use of it?” Which will come first, the chicken or the egg? Our brief report moves forward with the assumption that we must build rich authority databases Towards the end of our talk, we’ll pose a few persuasive arguments for the value of rich authority data and services that might arise from that enrichment endeavor
5. Jennifer: In thinking about how to enhance the discovery experience, we suggested policy changes that fall into two general categories: Changes that promote robust, granular data practices Changes that focus on adding meaningful linkages to our authority data Carl will walk us through a few examples
6. Carl: LC Name Authority File isn’t the only game in town. VIAF, ISNI and ORCiD are just a few examples of curated and self-registered authority databases.
7. Carl: The MARC Authorities Format makes it possible to include non-LCNAF identifiers in our NACO authority records, in the 024 field.
8. Carl: The subfield 0, as it is now, could be added to 5XX fields in an authority record.
9. Carl: Substituting subfields that are dedicated to specifying the source – subfield 2 – which is already in use elsewhere in authority records.
10. Carl: The parenthetical source notation in the subfield 0 of 5xx fields interferes with machine processing.
11. Carl: The 373 fields for Associated group indicate her affiliations.  Note the subfields 2, specifying that this is an authorized access point in an LCNAF authority record.
12. Carl: It is possible to find other sisters of the same order by searching LCNAF with Entity Attributes (via OCLC Connexion).
13. Carl: Need a citation for display purposes? Machines can follow those identifiers to the relevant bib records and assemble a MLA, APA, or Chicago style citation; when data is as rich as ours, we do not have to do all of the heavy lifting!
14. Carl: In this bibliographic record, the German National Library has included a subfield 0 containing an identifier in the 100 field.
15. Carl: View of the underlying data for the cataloging interface.
16. Carl: Cataloging systems can accommodate the world’s alphabets and scripts. With script and language data encoded in authority records, we might customize display based upon IP address, web browser setting, or a toggle in the discovery interface.
17. Jennifer: MARC isn’t the only game in town. The space we occupy is populated by many well established and up-and-coming metadata standards.
18. Jennifer: At Indiana, non-MARC metadata projects generate A LOT of authority work. More than we can keep up with. Working with visual and archival materials in particular provides ample opportunities for authority data creation and enrichment.
19. Jennifer: Frank Hohenberger Photograph Collection includes digitized documentary-style photographs from first half of the 20th century. In many instances, we find great contextual information about the contents of the photograph that we cannot actually include in the bibliographic record for the digital object.
20. Jennifer: This photo is of the medical offices of Ephraim McDowell, who in 1809 became the first man to remove an ovarian tumor in this very building in Danville, KY
21. Jennifer: Can I put the LCSH term “Ovariotomy” in the bib record for this photograph? No. The photo is about a building.
22. Jennifer: Much of the contextual data we learn in the course of describing archival collections belongs in authority data.
23. Jennifer: Imagine you’re on a reference desk. A high school sophomore approaches with a writing assignment and you must help her refine her research topic. What kinds of questions is it possible to answer with authority data alone? WWII fighter pilots who served in a particular squadron Baseball players born in New Jersey
24. Jennifer: 1. Search the rich authority data 2. Find relevant results 3. Click on a result and be taken to a landing page, which might look something like WorldCat Identities
25. Jennifer: 4. From there, an information seeker can go anywhere! Because works, creators, and topical subjects are linked via semantic relationships.
26. Jennifer: Information sourced from bibliographic or authority data may be composed on the fly and tailored to the user. Think of it as metadata cocktail made to order, presenting different information views.
27. Jennifer: By making authority data more linkable and granular: We can make use of authority data now! Big search engines have already abandoned keyword search algorithms in favor of semantic search. Our authority data helps drive semantic search forward We can publish our rich, valuable data (via APIs, etc.) so that others can make use of it We’ll be well-positioned for a data conversion to BIBFRAME or other linked data implementations We want better discovery interfaces for our users. Our group would argue that a discovery layer could have the best user experience design imaginable, without rich data under the hood, that discovery portal will be of little value
28. Indiana University PCC Non-MARC Authorities Issues Group. (2013 November 1). Brief on the Future of Non-MARC Authority. Retrieved from http://hdl.handle.net/2022/18401