BIBFRAME and Moving Away From MARC

Editor's Notes

1. I'm going to look at the: Background The model Some of the issues as we go along. Some other ways of doing it!
2. Short for BIBliographic FRAMEwork initiative Not an acronym! Also, Bibflow LibHub MarcNext
3. I alluded earlier to the problems with MARC and RDA in particular that LC and US testers identified… "The library community's data carrier, MARC, is based on forty-year old techniques for data management and is out of step with programming styles of today. No community other than the library community uses this record format, severely compromising its utility to other communities as a data transmission tool. Bibliographic applications being developed outside of the library environment are not making use of, and may not be compatible with, records encoded in MARC. New and anticipated uses of bibliographic data require a format that will accommodate and distinguish expert-, automated-, and user-generated metadata, including annotations (reviews, comments) and usage data. Flexible design should allow for the selective (modular) use of metadata in different environments (e.g., use of controlled vocabularies appropriate to specific domains). The existing Z39.2/MARC “stack” is not an appropriate starting place for a new bibliographic data carrier because of the limitations placed upon it by the formats of the past."—On the Record. Jan. 2008 "Many survey respondents expressed doubt that RDA changes would yield significant benefits without a change to the underlying MARC carrier. "Most felt any benefits of RDA would be largely unrealized in a MARC environment.  MARC may hinder the separation of elements and ability to use URIs in a linked data environment.  "Demonstrate credible progress towards a replacement for MARC" -- Report and Recommendations of the U.S. RDA Test Coordinating Committee. Executive Summary –June 2011 "Bibliographic framework is intended to indicate an environment rather than a "format"" according to the A Bibliographic Framework for the Digital Age (October 2011). However, most people when they are talking about Bibframe are in fact talking about this idea, which is what I'm going to focus on.
4. Storage - Possibly Manipulation - Yes Display – No (beyond actual data). Not like AACR2/RDA and MARC do. Input – MARC is basically an input screen in its own right. Bibframe editor could have been done years ago. Exchange and distribution – Yes, on the web Publication – Yes, on the web "Lingua franca of library cataloguing" – Hopefully not! MARC has a lot more purposes that it needs. This makes it hard for Bibframe if it tries to Replace MARC.
5. "foundation for the future of bibliographic description that happens on the web and in the networked world" AMBITIOUS! Is it too ambitious?
6. Conversion of MARC or replacement of MARC? – See the implications of the previous slide. Many purposes. For all libraries everywhere? For LC or for all academic libraries? For national libraries. Your local public library, school libraries? For all bibliographic data? – For repositories, digital image libraries, commercial data, Library Thing, article databases, discovery system knowledge bases, etc, etc. For books, AV, serials, anything held by a library?? On what system: ILSs, as an additional nicety, something else entirely? Is this all wise?
7. BIBFRAME is fundamentally "an initiative of the Library of Congress". Zepheira are consultants with expertise and experience with the Semantic Web. Worked with OCLC on schema.org as well as a new project with University of California, Davis to "investigate the future of research library operations, particularly the production of metadata — or data on data — and deployment on the Web." Its president, Eric Miller, has been prominent in the development of the Semantic Web and RDF. Interestingly not a formal collaborative effort. Zepheira are no longer formally involved although they have a number of related projects, including training. The emphasis in 2015 has been on keeping the format stable and encouraging libraries to test. I understand that something akin to the MARC management structure is likely at some point.
8. This is the basic Bibframe model. Bits of it look like FRBR, but the Expression in particular is missing. The Sir Humphrey version: "BIBFRAME has worked on modelling works as Works within the BIBFRAME model, similar to the RDA modelling work, itself modelled on the work on the FRBR model of Works and Expressions. A BIBFRAME Work is a creative work, perhaps a FRBR Work, or an RDA FRBR Work but it also expresses a FRBR Expression, and of course an RDA FRBR Expression. A Work may express another Work based on others’ work, not just a FRBR Work or an RDA Work. That also works. FRBR Works or RDA Works expressed as BIBFRAME Works can relate to FRBR Expressions (BIBFRAME Works or RDA Expressions). So, Works are works that can be Works but also Expressions linked to Works that really are Works."
9. Also similar to schema.org Work/? Split Schema.org has vaguely similar Work/Instance split. Oslo PL has similar Work/Manifestation split. The LC A/V Bibframe Modelling report suggested something different, creating an (unintentional/natural) Event category on the same level as Work and placing them both within something called Content. This report also has references to a lot of models that are not strictly FRBR or Bibframe or …
10. A Bibframe resource can be anything the same as as RDF resource is anything that can be given a URI. authorizedAccessPoint: "Controlled string form of a resource label intended to help uniquely identify it, such as a unique title or a unique name plus title." For the resource, not one of its access points. [Literal] identifier: Number or code that uniquely identifies an entity. [identifier] label: "Text string expressing the property value." [Literal] relatedTo: "Any relationship between resources." [URL]
11. classificationLcc: Other classifications are available. contains: "Work that is a discrete component of a larger work." creator: Creator role / Generalized creative responsibility role. hasDerivative: "Work has a modification for which it is the source. Work that has been translated, i.e., the text expressed in a language different from that of the original workp)." NB. expressionOf and hasExpression! hasInstance: "Work has a related Instance/manifestation."! bf:title: a string WorkTitle: a Title object
12. contributor: Contributor role / Generalized expressive responsibility role. dimensions: "Measurements of the carrier or carriers and/or the container of a resource." Isbn, isbn10, isbn13: More specific than bf:identifier. Identifier is a specific type of resource. identifierAssigner Identifier assigner. Literal identifierQualifier Identifier qualifier. Literal identifierScheme Identifier scheme. URL identifierStatus Identifier status. Literal identifierValue Identifier value. Literal. bf:publication: Seems to have replaced the previous attributes bf:placePub, bf:provider, bf:pubDate to provide a more event-based model (like the BL's). See example on next slide. titleStatement: "Title transcribed from an instance." For FRBR, note instanceOf. Also, at least in examples, the split between creator for work, and contributor for instances.
13. This is a brief example to give you an idea of the model as a whole, including Publication modelled as an event.
14. Example of Bibframe Authority. Previous properties included resourceRole, isnni, orcid, and viaf. Resourcerole is tricky.
15. EXAMPLE ISSUE 1: AUTHORITIES There was some debate on this issue, of how they should be handled. The first is what most people would assume would happen and is the classic linked data approach. Indeed, it's basically a recreation of the first example I showed you this morning. The second is Bibframe's attempt to deal with some essentially practical problems: Local indexing Local variation Allows augmentation from other sources Deals with people not in authority files without a pre-existing URI Although many of these things are possible by simply making use of existing techniques. There is also the danger that the notion of an authority is confused with that of a person. They are not the same thing
16. Which of these is it? There is evidence to suggest that it could be several. Person page says: "Individual or identity established by an individual (either alone or in collaboration with one or more other individuals)" but… Authority says: "Representation of a key concept or thing. Works and Instances, for example, have defined relationships to these concepts and things." There are also properties hinting at ambiguity or the traditional approach. authorityAssigner Authority assigner / Entity that assigned the information. Agent authoritySource Authority source / Authority list from which a value is taken. Literal hasAuthority Authority information / Link to controlled form of name or subject and other information about.
17. When knowing WHO is asserting the additional information is important!!! There is other work which approaches the same problems: named graphs, and an already established Open Annotations Model from the W3C whose outline Bibframe follows but not exactly. This could prove troublesome. Mooted as a possible way of modelling Holdings and items, but not without criticism.
18. bf:describes is a subproperty of bf:annotates bf:annotationAssertedBy: a URI of whoever asserted this was the case. The important bit for Annotations! In this case, UCL
19. This is an LC MARC record for FRBR: a guide for the perplexed by Robert L. Maxwell.
20. This is the same book but in Turtle. Difficult to fit on the screen!! I'll show you the whole thing, but split up and highlight a few things as we go along.
21. Note the LC URI for the Work at the top. No link to an external authority here, presumably because it's generated on the fly as a conversion.
22. There are two Annotations here (look for the square brackets). The first is a link to a table of contents. The second appears to be information about the creation of the record.
23. Instance, so has all the physical description, instance title, etc. Note the ambiguity of the bf:derivedFrom statement which implies that the the Instance is derived from the a MARC XML record.
24. Publication information is here.
25. One thing you may have noticed is that all the properties are BF specific…
26. EXAMPLE ISSUE 3: VOCABULARIES I've coloured in each line of a Bibframe "record" and you can see that most of it is red, which is the BF vocab. The orange is RDF The blue is MADS We heard about the BNB earlier and this is the same book…
27. The red here are the terms coined by the BL. All the others are external vocabularies.
28. I have highlighted the equivalents of the 100 and 245
29. This is The European Library. They recently converted a huge amount of RLUK data to linked data. The red here is in fact the RDA element set! The white is the EDM, and you can see it never made it beyond the prefixes.
30. The red here are the PL's own terms. OPL are working on producing and using LD like this for their internal purposes as MARC not fit for their purposes and they felt they couldn't wait for Bibframe to be ready. Do read the article! This and OCLC in particular point an intriguing way towards not waiting for a vendor to along with something ready for us to use.
31. Note here that The Title Proper is from the DCT vocabulary The Subtitle is from the Fabio vocabulary The SOR is from the RDA vocabulary! No information about the Creator here.
32. You can have a look at your own examples!!!