BIBFRAME, Linked data, RDA
6 likes•661 views
Robin Fay presented an update on the Bibliographic Framework Initiative (BIBFRAME). The presentation covered the need for BIBFRAME as MARC records have limitations for machine processing. FRBR and RDA were discussed as models that focus on relationships between works, expressions, manifestations and items. XML was presented as a way to encode bibliographic data in a machine-readable format using elements rather than character strings. The semantic web and linked data were discussed as ways to make metadata shareable on the web. BIBFRAME was introduced as a new bibliographic framework to replace MARC that would use RDF to encode bibliographic data.
![Lingering impact of the card catalog
Transition from card catalogs to electronic form of the card catalog
[MARC record] - duplication into an electronic format
“Space” an issue
• Need to save physical space on cards to reduce printing;
catalogers use abbreviations, etc.
• Early electronic records impacted by record and field limits –
remember flat files? Overwrites the whole file with a change!
• Current ILS are relational with field limits disappearing but
MARC records bibliographic length is 99999 characters
• Repeatable/split fields if needed, e.g., note fields; reliance on
abbreviations, conserve data
Focus on physical things - barcodes, circulation/patron records, starting
to shift as libraries have collected more digital items](https://image.slidesharecdn.com/bibupdate2016-190217233634/85/BIBFRAME-Linked-data-RDA-6-320.jpg)
![Instead of transcribing data using characters to denote
differences between types of data, clear & unique elements
can be used to differentiate them
264_1 … $c 1966 <publication date>1966</publication
date>
264_4 … $c ©1966 <copyright date>1966</copyright date>
264_1 … $c [1966?] <approximate
date>1966</approximate date>
Bibliographic data in XML](https://image.slidesharecdn.com/bibupdate2016-190217233634/85/BIBFRAME-Linked-data-RDA-23-320.jpg)
BIBFRAME, Linked data, RDA
- 1. Update on Bibliographic Framework Robin Fay, ATCx3 Consortia, Digital Initiatives @georgiawebgurl
- 2. Agenda The need for Bibframe FRBR /RDA Extensible Markup Language (XML) Semantic Web – metadata is everywhere Linked Data Bibliographic Framework Initiative (BIBFRAME) Discussion
- 3. Getting to know you Copy cataloging? Original cataloging? Catalog content or manage digital libraries, home grown databases (nonMARC), institutional repositories? Other technical services work (ordering, maintenance, authority work, etc.)? Runs system reports or bibliographic maintenance reports? Familiar with WEMI? Familiar with XML or other programming languages? Knows what a triple (or triplet is)? Can code MARC in their sleep?
- 4. Our history (MARC) MARC is considered a Resource Description Standard Developed at the Library of Congress during the 1960s by Henriette Avram Originally used as a way for LC to disseminate and print cards more easily & quickly MARC records largely look like electronic cards and function that way as well (more difficult for machines to process “strings” of data) Designed more for human-readability than machine-readability ISBD punctuation sometimes used to determine the meaning of a subfield; indicators often used to provide meaning (e.g., 245 – Title/statement of responsibility)
- 5. MARC, Indicators, subfields, ISBD & text strings – oh my! 245 10 Calm energy : ‡b how people regulate mood with food and exercise / ‡c Robert E. Thayer. MARC Tag Delimiter 2nd indicator 1st indicator Tags represent textual names They’re divided by hundreds: e.g., 100, etc. We haven’t used everything we could 000- 999 Indicators communicate information to the system What do 1st and 2nd indicator communicate?Robin Fay, 2010 ISBD ISBD (International Standard Bibliographic Description): Standardized punctuation (colons, semicolons, slashes, dashes, commas, and periods) is used to identify and separate the elements and areas.
- 6. Lingering impact of the card catalog Transition from card catalogs to electronic form of the card catalog [MARC record] - duplication into an electronic format “Space” an issue • Need to save physical space on cards to reduce printing; catalogers use abbreviations, etc. • Early electronic records impacted by record and field limits – remember flat files? Overwrites the whole file with a change! • Current ILS are relational with field limits disappearing but MARC records bibliographic length is 99999 characters • Repeatable/split fields if needed, e.g., note fields; reliance on abbreviations, conserve data Focus on physical things - barcodes, circulation/patron records, starting to shift as libraries have collected more digital items
- 7. Why now-Why RDA? Why FRBR? The world has changed since we implemented AACR2 – mobile, social, dynamic, online ! More flexibility in cataloging – more options Designed with digital items in mind including those in museums and archives (not just library-centric) More collaborative cataloging – harvesting from PDFs or other sources, outsourcing, sharing with each other (Transcription – should make cataloging easier) Considers how users use information – ability to bring together different formats and versions of the same work together more easily Expands concept of authorship to creator to encompass more roles – we are all content creators
- 8. Why now-Why RDA? Why FRBR? Looking forward to new systems – modern ILS are relational databases – clusters of tables with fields that can be linked in different ways Less reliance on flat file structures –more data in relational databases with minimal drain on resources More keyword and fulltext searching, which uses words in relevance ranking Breaking apart data into defined concrete bits allows machines to “think” and build content on the fly ; each word in fulltext (minus stop words – articles, etc.) is typically included with relevance weighting – the more it appears, the more relevant
- 9. Where the ILS is going… More recently ILS software is built on relational databases More user functionality - ability to share via social media, make “book bags” or reading lists Ability to work with other types of metadata Continuing to move away from the card catalog format Focus on authorities and controlled vocabularies - free text keyword searching newer functionality Moving more towards relational databases (remember FRBR) We need to move forward to> thus FRBR and RDA.
- 10. ER (Entity Relationship) Database an example
- 11. FRBR (Functional Requirements for Bibliographic Records) The FRBR report itself includes a description of the conceptual model of the bibliographic universe: that is, the entities, relationships, and attributes (or as we’d call them today, the metadata) associated with each of the entities and relationships, and it proposes a national level bibliographic record for all of the various types of materials. It also reminds us of user tasks associated with the bibliographic resources described in catalogs, bibliographies, and other bibliographic tools. -- Barbara Tillet, 2003 http://www.loc.gov/catdir/cpso/frbreng.pdf
- 12. FRBR Entities & Relationships Group 1: RDA* Group 2: FRAD Group 3: FRSAD
- 14. RDA RDA provides more flexibility in describing content, especially digital content. RDA emphasizes relationships, transcription, and more data not less*. While we typically use it in MARC, RDA can be expressed in XML or RDF (a semantic web framework). (We’ll look at RDA examples in XML in a few) We try to make MARC work for us – by creating machine actionable fields and $e and $i for building relationships
- 15. From University of North Texas Libraries’ catalog Copyright date is a separate element in RDA Role explicitly stated Relationship explicitly stated “pages” spelled out, not abbreviated p.
- 16. Semantic web is about data Big data - exactly what it says – large volumes or groups of data Linked data – a link that connects data together (relationships!) – it can be used to create dynamic displays (we already have linked data behavior with our authorized access points but no true linked data). 100s link to Authority records which then can do things like pull together a list of materials by an author Open data is data that is published with rights encouraging usage and sharing Social web
- 17. Allows us to have a customized experience on the web using any device that has data and internet capabilities (smart phones, tablets, laptops, ipods, desktops, etc.) It allows us to have better search results – personalized, with better relevance and filtering. It works for us. Semantic web – the user experience (UX)
- 18. Internet of Things – rise of connected machines and data sharing – Web 4.0?
- 19. • Types of metadata: • Descriptive • Structural • Administrative • Many forms of metadata include elements of each of these; however it is dependent upon the schema. • A schema is a set of rules covering the elements and requirements for coding. Examples of common schemas in the library world include Dublin Core/DCMI, AD, and others. Examples of schemas in the semantic web include Dublin Core/DCMI, FOAF (Friend of a Friend), and many others. Let’s talk about metadata on the web in general
- 20. Metadata in the wild (and some libraries) Structure – navigation, files/pages included, etc Description – A typical catalog record also information created by users Administration/administrative –who created it, how, file type, rights will sometimes fall here – DRM (Digital Rights Management), Provenance how you became the owner/repository for the item
- 21. Descriptive metadata Administrative metadata Descriptive metadata Descriptive metadata Digital libraries & metadata Administrative metadata
- 22. XML https://docs.google.com/spreadsheets/d/1tYxaiu28k4upbIQVbtNoWIkDfiZq3OuYeOJ2WM88emk/edit# gid=1557849273 Libraries use XML in a variety of ways. It can be used to crosswalk MARC into digital libraries (or the reverse!) Tools like MarcBreaker and MarcEdit Crosswalk can be used. A new version of MarcEdit is being released as MarcNext. Some digital library software have built in tools. A schema crosswalk shows how fields will be mapped from one to another.
- 23. Instead of transcribing data using characters to denote differences between types of data, clear & unique elements can be used to differentiate them 264_1 … $c 1966 <publication date>1966</publication date> 264_4 … $c ©1966 <copyright date>1966</copyright date> 264_1 … $c [1966?] <approximate date>1966</approximate date> Bibliographic data in XML
- 25. Make your own metadata – RIMMF for training Free software to create records for training (RDA in Many Metadata formats) using a form RDA to RDF, RDA to XML and even MARC ! http://www.marcof quality.com/wiki/ri mmf3/
- 26. What do you see as the challenges of thinking about bibliographic data in this way?
- 27. Semantic web & the social web So, what about the “rest” of the web? The social web/social media/social networking is about people. Its focus has been less on standards, controlled vocabularies, RULES…. After all, here comes everybody… ..but that is not exactly true. In order for blog posts to display sequentially, in order for discussions to be threaded, there must be an underlying order – a structure -- rules. Much of the technology (databases, servers, etc.) already exists – at least for the initial stages of the semantic web. Metadata schemas, RDF, are being used by many projects, but the web has only “standards” no rules.
- 28. Many terms associated with the semantic web are used or based upon information architecture, database, information science, and library science fields – controlled vocabularies, structural elements, etc. •RDF = Resource Description Framework •RDFS = Resource Description Framework Schema •OWL = Web Ontology Language – links ontologies which are classification systems •URI = Uniform Resource Identifier – we use these already A little semantic web terminology
- 29. The World Wide Web Consortium (W3C) standard, a simple metadata data model, based on triples (subject-predicate-object) ; developed in 1999, with RF 1.0 specification published in 2004. (1.1, 2014) It reflects the relationship of the items. (in other words, an entity-relationship model). It is meant to be neutral – vendor neutral and operating system independent. RDF is one way we could express bibliographic data in the future. We share commonalities between FRBR and RDF already. Examples The University of Georgia (resource as subject) is located in (predicate) Athens, Georgia(object). The Raven is written by Edgar Allan Poe. RDF: Resource Description Framework Typically it is coded in XML.
- 30. Some RDA RDF properties Identifier for the work Date of work Language of expression Media type Carrier type Variant title Preferred name for the person Date of birth Variant name for the corporate body RDA expressed as RDF in XML – likely!
- 31. The Semantic Web is based upon more precise utilization of data and is heavily dependent upon The code The metadata and its metadata schemas (rules) The ability for machines (including devices and home appliances) to talk to each other and make sense of that communication Linking data makes this process easier since we do not have to re-enter data, we can just link to it. This can work for libraries, reducing input and also maintenance. Linked data may help solve issues
- 32. Linked data Linked data is about connecting data and reusing data – rather than see text in a record, we would just see a link Linked data is: “about using the Web to connect related data that wasn't previously linked, or using the Web to lower the barriers to linking data.” Benefits – maintenance work! Authority work! Less keying (possibly!)
- 34. The thesaurus consists of more than a million terms organized into five controlled vocabularies: subjects, personal names, organizations, geographic locations and the titles of creative works (books, movies, plays, etc).” – NYT Blogs Linked data may help solve issues Projects such as the NYT Linked Open Data project and the Virtual Authority File project are resources of controlled vocabularies. LCSH has been released as linked data too! http://id.loc.gov/
- 35. Bibliographic Framework Initiative “…aims to re-envision and, in the long run, implement a new bibliographic environment for libraries that makes "the network" central and makes interconnectedness commonplace.” (read: is attempting to better position the library world for a linked data environment) Primer for BIBFRAME: http://www.loc.gov/bibframe/pdf/marcld-report-11-21- 2012.pdf
- 36. Library of Congress Consultants: •Zepheira Partners (Early Experimenters) and among others: •British Library, •Deutsche National bibliothek, •George Washington University, •National Library of Medicine, •OCLC, •and Princeton University Initial participants
- 37. Officially launched by the Library of Congress in 2011 A new model for bibliographic data, that will be the basis for an new encoding standard that will replace MARC and will be XML-based. Consists of the BIBFRAME Model is a conceptual/practical model that contains 4 high-level classes, or entities (Work, Instance, Authority, and Annotation) and the BIBFRAME Vocabulary which has a defined set of elements and attributes that describe resources and their properties. BIBFrame: BIBliographic FRAMEwork Initiative
- 38. Instead of bundling everything neatly as a “record” and potentially duplicating information across multiple records, the BIBFRAME Model relies heavily on relationships between resources (Work-to-Work relationships; Work-to-Instance relationships; Work-to-Authority relationships). It manages this by using controlled identifiers for things (people, places, languages, etc). MARC employs some of these ideas already (geographic codes, language codes) but BIBFRAME seeks to make these aspects the norm rather than the exception. In short, the BIBFRAME Model is the library community’s formal entry point for becoming part of a much larger web of data, where the links between things are paramount. (from BIBFRAME FAQs: http://www.loc.gov/bibframe/faqs/) BIBFRAME
- 39. BIBFRAME Model Core elements of the BIBFRAME model – similar but not exactly the same as FRBR WEMI but maps to WEMI Work resource reflecting the conceptual essence of the cataloged item Instance Resource reflecting a material embodiment of a BIBFRAME work Authority Resource reflecting key authority concepts that have defined relationships to
- 41. Links replace strings = Reduced maintenance URIs = authority 264 1 New York: Neal Schuman Publishers, 2014 Instead of writing New York, replace it with http://id.loc.gov.authorities/names/n92062246 Replace Neal Schuman Publishers with http://id.loc.gov/authorities/names/n79007751 264 example
- 42. The model
- 43. Relationships
- 46. BIBFRAME Primer (2012) - http://www.loc.gov/bibframe/pdf/marcld-report-11-21-2012.pdf Work
- 47. Instance
- 48. 49 http://www.loc.gov/bibframe/docs/vocab-conventions.html The BIBFRAME Vocabulary is comprised of the RDF properties, classes, and relationships between and among them.
- 49. http://bibframe.org/vocab-list/ List of classes (types) from Bibframe.org Each is further defined with a description and usage.
- 50. Similar to DublinCore / DCMI’s structure on the web.
- 51. http://bibframe.org/vocab-list/ 52 List of properties (relationships)
- 52. Naming conventions for classes and properties http://www.loc.gov/bibframe/docs/vocab-conventions.html
- 56. Examples – BibFrame in MARC
- 59. Examples – Creating records
- 63. 64
- 65. Click on “Save” at the bottom of the page to generate a BIBFRAME view
- 67. Will RDA elements be part of the BIBFRAME vocabulary? Yes. RDA is an important source of elements in the vocabulary for BIBFRAME, even though it generally aims to be independent of any particular set of cataloging rules. We also expect community profiles to emerge which will accommodate additional elements. (from BIBFRAME FAQs: http://www.loc.gov/bibframe/faqs/)
- 68. BIBFRAME bf:tableOfContents = AACR2 1.7B18 – Contents = RDA 25.1 – Related Work = MARC 505 – Formatted Contents Note
- 69. And back to MARC…. We can’t abandon MARC immediately No current viable replacement that works with current ILSes (but getting closer with Bibframe!) we have MILLIONS of records in MARC (need conversion tools or systems that easily support multi metadata schemas – displaying materials equally in an understandable way to users >“mapping”) We need to change the way we think about cataloging (FRBR) to a lesser extent how we implement cataloging (RDA) – print is no longer the default material format and more importantly the metadata that is out there on the web. ╫
- 70. 71 Lots of Bibframe resources & links https://goo.gl/pih0Ui [email protected]