CDI to manage OA resources (Implementation+Post-live)

Legend: not started IN PROGRESS STALLED decided

Recommendation

Implementation phase

After discussing with RMFG and SCLG E-book Strategies Team, PSELG’s firm recommendation is that all campuses migrate all OA records that have been cataloged locally. We are unable to harmonize policy before implementation and go-live.

Testing and After Go-Live phase 

PSELG recommends the immediate formation of a cross-functional temporary subteam (composed of representatives from scholarly communications, technical services, and public services domains) to investigate how best to manage the OA resource activation in the Test Environment CDI to give local implementation teams and CDL staff sufficient time to guide decisions impacting go-live and continuing operations. 

PSELG further recommends that this subteam conduct a review of our current Policies and Procedures for Shared Cataloging, Linking, and Management, make recommendations to JSC and/or SCLG for updates to the Policies and Procedures, and develop a systemwide standard practice of how and when OA resources are included in CDI. This cross-functional subteam should continue past go-live to complete its responsibilities.

Background

Summary

In November 2020, PSELG was asked by PPC to address the question: “Should we, as a system, use CDI to manage OA resources?” The question came up through the SCLG E-book Strategies Team, but similar discussions also occurred in the Technical Services Escalation Leads group (TSELG) and other functional groups before being brought to PPC.  While simple on its face, the question has multiple implications for UC-wide policy on resource management, individual campuses' workflows, and shared services like the Shared Cataloging Program (SCP). PSELG led foundational interviews/discussions with representatives from TSELG, the Joint Steering Committee on Shared Collections (JSC), the Resource Management Functional Group,  and interested campus parties about the initial question and implications for systemwide and campus workflow. Each group was given a brief summary of the question and its history and then asked for expert opinions related to their area of knowledge. Campus groups were also asked, “Is your UC campus still committed to the Policies and Procedures for Shared Cataloging, Linking, and Management?  (yes / yes with revisions / no).” Six out of ten UC campuses (Berkeley, Irvine, Los Angeles, Santa Barbara, Santa Cruz, and San Francisco) answered: “Yes, with revisions.” Merced said they were committed to the current policy.  Two campuses (Davis and San Diego) expressed a variety of differing opinions; however, more information was needed to determine. One campus (Riverside) felt PSELG was not the appropriate venue for this discussion. It was agreed upon by PSELG and those interviewed that this is not a question that needs to be resolved in the Vanguard phase. However, the question does need to be addressed by a cross-functional team of campus and stakeholder groups in the SILS process as it generates many related questions that impact a variety of levels at UC Libraries, including but not limited to 1) philosophical questions about the purpose of discovery layers, 2) practical questions about procedural gains among libraries in a shared ILS experience, and 3) pragmatic questions about user experience in discovery. The following decision page documents the recommendation of PSELG and outlines questions for consideration beyond the purview of PSELG.

The summary of questions below in the “Questions to Consider” section illustrates the interdependencies of several SILS functional groups and Common Knowledge Group areas of expertise in the UC Library structure.  While some questions are practical questions that will be addressed as we move further into implementation, the underlying set of challenges expose a lack of common understanding of next-generation discovery principles in the intersection of the operational theory of technical service standards (e.g., cataloging, indexing, acquisitions) and the user experience layer of the search product. These questions have an interest to a broad set of functional groups (Acquisition & E-Resources, Digital Collections, Discovery, Resource Management), all three escalation groups, and both Training & Outreach groups.

Reference note: The Central Discovery Index (CDI) is Ex Libris’ latest index for 4+ billion resources, including content from databases, digital repositories, open access content, and more. The goal of CDI is to offer users access to a wide range of resources with enhanced search capabilities.

I. Findings related to the philosophy of discovery layer 

The Discovery Functional Group has developed a vision for the SILS discovery layer and also continues to document configuration and harmonization recommendations based on this vision and usability testing results. The vision document provides an overarching philosophy and guide for design, usability, harmonization, and continued configuration/iteration. This higher-level work is helpful when thinking about CDI & OA’s impact on discovery.

PSELG’s discussions around CDI & OA within discovery are aligned with the Discovery FG’s vision. We agree that OA content is important for users to discover but requires careful curation and high maintenance upkeep. When not managed well, searches could result in high confusion and frustration for users.

Related Documents:

• https://docs.google.com/document/d/1VtV2Q1vV3cuXThk-iL7tp8-_U0i1bxcpYTM7E2Pknuo/edit?usp=sharing  

• https://uc-sils.atlassian.net/l/c/JV6dz5Ve (Decision Page) - Decisions being made during the Vanguard and Testing phases of SILS incorporate the CDI. These decisions are helpful to be aware of when considering the next decisions about OA materials in the CDI. 

II. Findings from discussions with JSC and other groups: UC OA cataloging policy

A. MEETINGS WITH SILS/UC-WIDE STAKEHOLDER GROUPS

The PSELG co-chairs, Rikke Ogawa and Beth Callahan, met with several different groups who have an interest in this topic. Groups/individuals are listed below with links to notes.

• Liz Miraglia, Chair of the Resource Management Functional Group
  Major takeaways: the issue of OA records is far-reaching and complex; we will need a way to test OA resource records in the future.

• Xiaoli Li and John Riemer, Technical Services Escalation Leads co-chairs
  Major takeaways: Areas for further exploration include campus experiences with OA resource discovery using CDI versus cataloging, SCP approval criteria for adding an OA resource, and the various campus approaches to cataloging OA resources. It would be useful to have a group to explore long-term decisions around cataloging policies and procedures, SCP priorities, and whether the implementation of CDI aligns with UC commitments to OA resources.

• Shi Deng, SCLG ebook Strategies Team representative
  Major takeaways: OA resources managed by CDL follow the UC OA resources policies and procedures. Many questions arose related to new features in Alma/Primo (e.g., CDI, metadata quality, designated responsibility of NZ-related tasks with UC structure, etc.).

• Ivy Anderson and Michael Walmsely (Joint Steering Committee (JSC), chair & member) led a discussion on the topic at the January 2021 JSC meeting. 
  Major takeaways: JSC doesn’t feel OA resources should be disfavored or have 2nd class status in relation to licensed resources. Evaluation of CDI records needs to be part of the process, similar to what we do now for cataloging of licensed resources. Collection development staff should be the ones to evaluate whether CDI resources are ‘good enough’ for OA discovery. 

Related Documents:

• Notes from meeting with Liz Miraglia (Chair, RFMG): https://docs.google.com/document/d/1ZmFvbmkt_wgB3BohzWekR6Q12paMQnJ6z5G2gW1X4B0/edit?usp=sharing

• Notes from meeting with Xiao Li & John Reimer (Co-chairs, TSELG):

• Notes from Meeting with Shi Deng (SCLG Ebook Strategies Team):

• Notes from meeting with Ivy Anderson & Michael Walmsely:

• JSC email response after discussion at January 2021 meeting:

B.  MEETING WITH CAMPUS STAKEHOLDERS

Campuses had a variety of opinions about pursuing the use of CDI to replace continued SCP and campus cataloging of Open Access resources. A common area of concern across campuses was the importance of both user experience and user expectations for discovery. Campuses who indicated they were no longer committed to the UC Policies and Procedures for Shared Cataloging, Linking, and Management in its current form provided the following opinions and insights:

• There is a clear value in having a common understanding and approach to cataloging resources.

• Campuses question the need to have one centralized point for cataloging, such as SCP.

• More than one campus commented that cataloging journals might no longer be a priority because articles are easily discoverable through other means. 

• Campuses felt that we should expend shared cataloging resources on cataloging monographs, which may be difficult for local campuses to accomplish.

• UC-wide tension exists about how we demonstrate commitment and value to the access of subscribed, purchased, or UC-created OA content (i.e., prioritizing cataloging/indexing titles where we have made investments with publishers in transformative agreements).

The campus discussions generated many questions that plumb the depth of CDI, cataloging, user experience, etc. We include a high-level set of summary questions from campuses and stakeholder groups in the Questions to Consider section listed below. For an extensive list of questions, please consult the Google Drive documentation linked directly under this paragraph in related documents.

Related Documents:

• Feedback from Key Stakeholders at each of the UC Libraries:

• UCLA document of expanded notes from campus discussion:

• Extensive list of questions from Campus discussions:    

III. Central Discovery Index

UC Libraries could activate OA resources in CDI for discoverability by users. However, decisions around activation of resources in CDI will require consultation with numerous groups to discuss the end-user experience and, if implemented, appropriate policies and viable workflows.

The variety of campus feedback offered by key stakeholders at each of the UC Libraries emphasizes the range of unanswered questions. Many of these questions are similar to questions raised by SILS/UC-wide stakeholder groups. For brevity, duplicate questions/ideas previously raised were removed from this section but continue in all sections' related documents (I-III). Many questions related to the lack of hands-on experience in Alma/Primo in a consortial setting and, at the time of discussion, no clear shared governance structure post-implementation. Highlights include:

• Questions about which other SILS groups are investigating this question, who will make implementation decisions, who will be responsible for the ongoing work, and what process would be used to change our approach later if we find the original vision does not work for us.

• Questions about the relationship between content in the CDI and UC discovery resources (such as eScholarship, Primo, and the Knowledge base), including which content is in the NZ, IZ, and CZ. 

• Questions about the scope of collections and how granular the selection of content we can turn on in the CDI: 

  • some campuses desire a high level of curation or, conversely, as a little review of individual resources as possible if it’s an agreed-on type of content such as from core publishers; 

  • some campuses desire to focus on those things we invest financial resources in and those things that are otherwise not discoverable.

• Questions about whether the designations of OA content in the CDI are accurate.

• Questions about how the CDI relates to our goals of reducing duplication of content and reducing unnecessary work.

• Suggestion to learn more from peers, such as Minnesota’s decisions about CDI and Penn State’s article about increasing OA resources' visibility in library catalogs.

Feedback from PPC on these questions indicates that some SILS groups are already at work on some of these issues (Acquisition & E-Resources Functional Group, Discovery Functional Group, and Holly Eggleston at CDL). Furthermore, it was suggested that RMFG and SCLG could be more formally involved in investigating and decision-making. Currently, it does not seem that any group is charged with coordinating among these various functional areas on policies and configuration. It was suggested that oversight for this issue might be split between a policy group and an operational group.

Related Documents:

• Questions from PSELG to PPC:  

IV. Further Resources for learning about CDI, UC policy, OA cataloging, etc.

• CDL page on OA Resource Cataloging: https://cdlib.org/services/collections/licensed/policy/open-access-resources-at-the-uc-libraries/

• CDI Collection List: Column H in the spreadsheet indicates Full-Text Rights with Open Access as one designation.

• Open Access Content in CDI (Knowledge Article): Notes that an Open Access designation point to primarily unrestricted content though a small portion may be restricted.

• Open Access Indication in Primo VE: Journal content appears to receive an Open Access indicator from Ulrich’s. Note: The article is referencing PCI rather than CDI. 

• ERM Panel Webinar (CDI focus) @ ELUNA California User Group site. See webinar and slides from CSU librarians (Magedanz, Hartwigsen, Holmes) regarding configuring/activating CDI collections (Nov. 2019).

Questions to consider

Internal Library- focused questions that were unclear to many sub-groups (PSELG, RMFG, TSELG, etc.):

• How does the CDI operate? What is the process for including a resource? 

• How are decisions made with regards to the content (including OA) that appears in the discovery layer? Are there some existing criteria for inclusion or exclusion of a resource? Will discovery layer inclusion/exclusion decisions be campus-specific? Is there a set of content inclusion decisions that will be made as a system?

• Our understanding is imperfect, though currently, it seems that we will not be able to see the underlying metadata from any CDI resource. What are the markers that will indicate the quality of metadata is good enough?

• With the implications of UC exploration of transformative collections and the focus on future “ownership” models of content, is there a common standard for a UC approach to inclusion/exclusion of resources that we invest service in but do not necessarily own?

User Experience / Discovery

• What is the desired user experience? For example, is the discovery layer's purpose of bringing the user a faceted set of results, mixing result sets between subscribed and owned library resources? Or is the purpose to provide a Google-like experience where the most relevant content displays first?

• What is the purpose of the discovery layer? Creating an intersection between catalog (e.g., book, series, etc.) and contents of the items cataloged (e.g., book chapter, journal article, etc.)?

• How will the content (including OA) display in discovery (look & feel of the user interface)?

• Which group(s) are charged with investigating impacts of configuration options, the scope of collections turned on, and addressing overall vision and implementation decisions for the CDI? 

• What is the anticipated user experience for OA collections activated in the CDI?

• Which groups best understand how OA activation in the CDI will relate to discoverability and access?

• Do the SCP full records (current policy is to catalog at the database/title level) allow for more or better discoverability than the CDI records?

 Local / UC-wide Decisions

• Will OA collections shared in the Network Zone (NZ) be inherited by all UC Libraries?

• Will there be agreed-upon content -- including OA collections -- to activate at this level? If so, who will be responsible for this work? If so, what activation configurations are best (e.g., using “link in the record”)? What level of work will still be required by local campuses?

• In what ways might this work overlap with resource management of shared collections, whether Tier 1 or Tier 2?  This could be considered in any OA / CDI workflow conversations.

• How might using the CDI for OA collections enhance, complement, or integrate with UC Libraries’ current practices for making OA materials discoverable?

• How will the implementation avoid unnecessary duplication of content in the displays?

See other decision pages:

Action Log