Enthusiasm growing for automated data sharing for knowledge base management

Librarians, publishers, and vendors are starting to embrace automated data exchange as a best practice. As this new approach gains acceptance, a good dialog is needed among all parties to ensure consistent approach across publishers and vendors, and realistic expectations among librarians.

Pictured here: Jozef Paulik, Senior Manager, Product Technology at Elsevier/ScienceDirect, speaking at ER&L 2016

Automated data exchange for knowledge base management is an idea whose time has come. The 2016 ER&L conference in April saw two sessions devoted to the topic that were well attended by librarians, publishers, and vendors. The interest and enthusiasm is understandable because automated data exchange has the potential to provide faster, better knowledge base management with far less effort. At ScienceDirect, we have had a system in place for more than a year that allows libraries to share their holdings data with the partners they choose. We now have 281 libraries that have automated data exchange with OCLC to update ScienceDirect holdings in their WorldCat knowledge base; 46 of these libraries use it to keep their WorldCat Local catalog up to date and 235 use it to automatically manage their MARC record deliveries for ScienceDirect books. Another 500 libraries have turned on data exchange for article linker, discovery, and catalog services through Ex Libris/Alma. Ebsco is currently working with Elsevier to implement data exchange as well. In this article, I’d like to share what we have learned so far.

It all starts with KBART.

Elsevier’s Jozef Paulik was working on improvement to the ScienceDirect electronic holdings reports system at the same time that the KBART II recommendation was being completed. The KBART II recommendation deals primarily with reporting collection titles, but it also covers custom title lists to match consortium purchases. Jozef realized that the recommendation for consortium reports could be applied to a holdings title lists for a single library. So ScienceDirect now allows you to download a list of your library’s holdings in KBART format (other formats including Excel are also available). The next step was to talk to service providers who were gearing up to ingest KBART collection reports from publishers to see if they would work with Elsevier to create a direct update of their library knowledge base using the KBART holdings reports we had developed. OCLC was the leader here, and was able to make automated holdings updating a feature of their new WorldShare Collection Manager.

The knowledge base has to match entitlement, not purchase history.

At Elsevier, we built our library reports to use our access and entitlement data, not customer sales history. This approach ensures that the knowledge base matches exactly what a user at the library can access on ScienceDirect. This way of thinking takes a bit of adjustment for libraries that are used to receiving a set of MARC records or selecting a collection in an admin tool to match a particular purchase. From a practical point of view, automating knowledge base management can only succeed if we limit the scope to the key function of the KB, namely, providing holdings data to support discovery and access to library content. This focus does not do very much for the acquisitions function of the library, and does not show how or when or under what terms a book or journal was purchased. Those necessary functions should also be supported by publishers with more automation, but it will require another set of tools and procedures.

Data exchange is a win-win-win-win.

The shortcoming of KBART II is that, by itself, it doesn’t save any effort for libraries. It standardizes the collection title lists that publishers exchange with vendors. The collections are more accurate and updated more quickly, but a librarian still has to place an order for MARC records or go in the admin tool of the article linker or discovery service and select collections to match their holdings. The inevitable delays in this manual selection process made more work for librarians, vendors, and publishers; and it undermines the accuracy and timeliness of knowledge base data, making it harder for researchers to find and access the online materials in their libraries. If we leave collections behind and move to title-level knowledge base management using automated data exchange, libraries win, publishers win, vendors win, and researchers win.

Success depends on collaboration.

While the KBART II recommendation provides a strong basis for automated exchange of holdings data, it was not written for that specific purpose. All parties will benefit from collaborative and open development of best practices now, when automated data exchange is poised to become widely adopted. Elsevier was pleased to work with representatives from Springer, JSTOR, Project Muse, and Wiley on the Success Strategies for Electronic Content Discovery and Access white paper that lays out the foundation for much of this work. We look forward to the next phase of collaboration, and have begun to have conversations with other publishers who share our enthusiasm for open development of best practices for holdings data exchange.

“The desire for automated data exchanges in the library community is undeniable. At ER&L, I attended two related sessions that were so heavily attended, later comers were relegated to standing room only. It is a publisher and vendor responsibility to help libraries maximize their workflow efficiency and minimizing their efforts. Automated data exchanges help accomplish exactly that. Wiley is investigating how to implement this practice and hopes to bring it to fruition.” –Matthew Ragucci, Library Technical Services Specialist at Wiley

The new approach is challenging but worth it.

Earlier this year, I conducted an informal survey that recruited 20 participants from the ALCTS listserv who have implemented automated data exchange at their libraries. The responses presented below show that the new technology has required new procedures and responsibilities, and that there have been challenges in getting it set up. Despite these challenges, there was an overwhelming endorsement when the survey asked these librarians wished that more publishers and aggregators would adopt automated data exchange.

Frequently Asked Questions about ScienceDirect MARC records and data exchange

The Elsevier.com website provides information ScienceDirect automated data exchange and MARC records program.

Q: Where can I find my ScienceDirect data exchange token for OCLC?
A: You can create your own sharing token in the ScienceDirect Admin Tool or we will create one for you if you contact the ScienceDirect Helpdesk.

Q: I have asked OCLC to enable data exchange. Which collection do I need to select in Collection Manager in order to receive MARC records for my ScienceDirect holdings?
A: Once OCLC has set up the data exchange with ScienceDirect, you should find the “ScienceDirect All Books” option has been automatically selected for your library and your holdings title count shown next to it. You should *not* select this collection manually. If you do not see the “ScienceDirect All Books” collection in your holdings on date when OCLC said they would complete your set up, please contact OCLC support .

Q: Why don’t I see MARC Records for my frontlist collections in Collection Manager?
A: To ensure accuracy, ScienceDirect does not release MARC records in advance. MARC records for new titles are distributed about two weeks after each title is published on ScienceDirect.

Q: Do I also need to activate MARC records for ScienceDirect Book Series in Collection Manager?
A: No. Book series volumes are included in the “ScienceDirect All Books” collection.

Q: I have set up MARC Records via Collection Manager, but how do I know I have set up automatic delivery? Will I still need to request records?
A: If you are unsure, please contact OCLC support to check whether you have enable automatic delivery for your ScienceDirect MARC records.

Q: We have just purchased additional books on ScienceDirect, do I need to order MARC records for them?
A: No. Once you have enabled data sharing and automated MARC delivery, you do not need to place an order for ScienceDirect MARC records again. Whenever books are added to your institutional holdings, either through a new purchase or when new titles are published on ScienceDirect, you will receive the corresponding MARC records.


Full results of the informal survey are posted at https://www.surveymonkey.com/results/SM-TZB2TQYR/. Individual comments have not been shared.

Kemperman, Suzanne Saskia, Bill Brembeck, Elizabeth W. Brown, Alexandra de Lange-van Oosten, Theodore Fons, Catherine Giffi, Noah Levin, Alistair Morrison, Carlen Ruschoff, Gregg A. Silvis, and Jabin White. 2014. Success Strategies for Electronic Content Discovery and Access: A Cross-Industry White Paper. Dublin, OH: OCLC.

Contributed by:

Alistair Morrison, ScienceDirect Product Manager for Books