Skip to content

Hanging Together

the OCLC Research blog

  • Home
  • About
Main Menu
Libraries / Metadata

Working in Shared Files

April 7, 2015May 11, 2017 - by Karen Smith-Yoshimura

Working in Shared Files blog graphic

That was the topic discussed recently by OCLC Research Library Partners metadata managers, initiated by John Riemer of UCLA. Working in shared files is a critical efficiency to free up time to address new metadata needs and roles. Metadata managers who need to allocate staff to cover more objects of interest to researchers in the information landscape and at the same time preserve metadata describing this material have every incentive to consider working collaboratively, in shared files.

Libraries have tended to treat WorldCat as a resource to be further edited locally. The 2009 report Study of the North American MARC Record Marketplace bemoaned the “widespread resistance to the idea of simply accepting the work of another library.”  We have been saddled with hundreds of copies of records across libraries and constrained to limit the amount of catalog maintenance done. When Kurt Groetsch described how Google was attempting to take advantage of library-created metadata during the 2010 ALA Midwinter meeting, he noted they “would like to find a way to get corrected records back into the library data ecosystem so that they don’t have to fix them again.” The linked data environment offers a new opportunity to create and maintain metadata only once and simply pointed to by all interested parties.

The discussions revolved around these themes:

Sharing edited records: In general, staff focus on only editing records that affect access points. Most libraries accept vendor records or records for shelf-ready books without review. Vendor records may need to be modified for the data to be consistent and linked. Vendor records are of varying quality, some of which hinder access. It was suggested that libraries can advocate vendors’ contracting the metadata creation with OCLC as part of their purchase negotiations. [Note: Focus group member Carlen Ruschoff of University of Maryland served on the cross-industry group that identified problems in the data quality in the content supply chain and gave practical recommendations for improved usage, discovery and access of e-content. See the 2014 OCLC white paper, Success Strategies for Electronic Content Discovery and Access.]

Policies and practices have been put in place to stop staff from doing what they don’t have to do. “Reuse rather than modify.” But it can be difficult to stop some staff from investing in correcting minor differences between AACR2 and RDA that don’t matter, such as pagination. One approach is to assign those staff important tasks (create metadata for a new digital collection for example) so that they just don’t have time to take on these minor tasks as well. Not everyone can accept records “as is”, but with all the effort the community has invested in common cataloging standards and practices, if we all “do it right the first time” we should be able to accept others’ records without review or editing.

When edits are applied to local system records, or other databases such as national union catalogs, the updated records are not contributed to WorldCat. The University of Auckland uses four databases: the local database, the New Zealand National Union Catalogue, WorldCat and the Alma “community zone” available only to other Ex Libris catalogers. When Library of Congress records are corrected in WorldCat, the corrections are not reflected in the LC database. When OCLC loads LC’s updated records, any changes that had been made in the WorldCat records are wiped out. We need to get better at synchronizing data with WorldCat. Perhaps updated “statements” can be shared more widely in a linked data environment?

Sharing data in centralized and distributed models: Discussants were divided whether a centralized file would be needed in a future linked data environment where WorldCat became a place where people could simply point to. Developers say there is no need for a centralized file; data could be distributed with peer-to-peer sharing. Others feel that a centralized file provides provenance, and thus confidence and trustworthiness. How would you be able to gauge trustworthiness if you don’t have that provenance pointing you to an authoritative source?

The OCLC Expert Community expanded the pool of labor able to make contributions to the WorldCat master records. This offers a new opportunity for focus group members who have been working primarily in their local systems. OCLC’s discontinuation of Institution Records is prompting some focus group members who have been using them to rethink their workflows, determine what data represents “common ground” and consider using WorldCat as the database of record. The OCLC WorldShare Metadata Collection Manager treats WorldCat records as a database of record and allows libraries to receive copies of changed records. It was noted that controlling WorldCat headings by linking to the authority file obviates the need for “authority laundering” by third-parties.

Importance of provenance: Certain sources are more trusted and give catalogers confidence in their accuracy. Libraries often have a list of “preferred sources” (also known as “white lists”.)  Some select sources based on the type of material that is being cataloged, for example, Oxford, Yale and Harvard were mentioned as a trusted source for copy cataloging old books on mathematics. Another criteria is to choose the WorldCat record which has the most holdings as source copy.

Sharing statements: Everyone welcomes the move to use identifiers instead of text strings. Identifiers could solve the problem of names appearing in documents harvested from the Web, electronic theses and dissertations, encoded archival aids, etc. not matching those used in catalog records and the authority file. Different statements might be correct in their own contexts; it would be up to the individual or library which one to use, based on what you want to present to your users. In a linked data world one can swap one set of identifiers with another set of identifiers if you want to make local changes. In the aggregate, there would be tolerance for “conflicting statements” which might represent different world views; at the local implementation level you may want to select the statements from your preferred sources. Librarians can share their expertise by establishing the relationships between and among statements from different sources.

Some consider creating identifiers for names as one of their highest priorities, spurred by the increased interest in Open Access. For researchers not represented in authority files, libraries have started considering implementing ORCIDs or ISNIs. [See the 2014 OCLC Research report, Registering Researchers in Authority Files.]

Karen Smith-Yoshimura

Karen Smith-Yoshimura, senior program officer,  topics related to creating and managing metadata with a focus on large research libraries and multilingual requirements. Karen retired from OCLC November 2020.

oclc.org/research/people/smith-yoshimura.html
Share on Facebook
Facebook
Tweet about this on Twitter
Twitter
Email this to someone
email

Related Posts

Visualizing Digital Humanities Workshop

July 10, 2017July 7, 2017

Post-its from workshop

Combating fake news through “credibility standards”

May 25, 2017April 29, 2019

A story about the spirituality of libraries

May 31, 2016May 31, 2016

OCLC Research

Hanging Together is the blog of OCLC Research. Learn more about OCLC Research on our website.

Links

  • Lorcan Dempsey's Weblog
  • Next – OCLC Blog
  • OCLC Research
  • OCLC Research Library Partnership
  • WebJunction

Categories

  • Archives and Special Collections (195)
  • Born-Digital Special Collections (14)
  • Collective Collections (118)
  • Data Science (7)
  • Digital Preservation (69)
  • Digitization (24)
  • Equity, Diversity, Inclusion (EDI) (7)
  • Evolving Scholarly Record (10)
  • Higher Education Future (8)
  • Identifiers (26)
  • Infrastructure and Standards Support (88)
  • Libraries (93)
  • Libraries Archives and Museums (125)
  • Libraries in the Enterprise (1)
  • Library Management (5)
  • Linked Data (33)
  • Measurement and Behaviors (44)
  • Metadata (75)
  • Miscellaneous (176)
  • Modeling new services (112)
  • MOOCs (7)
  • Museums (57)
  • Open Access (14)
  • Renovating Descriptive Practice (114)
  • Research Data Management (19)
  • Research Information Management (35)
  • Research Library Partnership (161)
  • Research support (22)
  • Resource Sharing (8)
  • Searching (38)
  • SHARES (6)
  • Supporting Scholarship (65)
  • Systemwide Organization (42)
  • User Behavior Studies and Synthesis (6)
  • Visual Resources (17)
  • Web Archiving (14)
  • WebJunction (6)
  • Wikimedia (43)

Share Buttons

Share on Facebook
Facebook
Tweet about this on Twitter
Twitter
Email this to someone
email

Email Notifications


 

Recent Comments

  • Gail Thornburg on さようなら (Sayōnara)
  • Ivy Anderson on さようなら (Sayōnara)
  • Günter on さようなら (Sayōnara)
  • Shuwen Cao on さようなら (Sayōnara)
  • Andrew Padilla on Presenting metadata from different sources in discovery layers

Recent Posts

  • Frequently asked questions: resource sharing practice in the time of COVID-19, Phase I
  • Towards respectful and inclusive description
  • The way forward to a more open future … together
  • さようなら (Sayōnara)
  • OCLC-LIBER Open Science Discussion on Citizen Science

Admin.

  • Log in
  • Entries feed
  • Comments feed
  • WordPress.org
  • [Un]Subscribe to Posts
© 2020 OCLC