Data Issues Task Force Open Forums

Suggested starting topics:

  • Proposal to ExL developers for better handling of Primo title-only portfolio matching, when the title is under 10 words  (issue widely seen in Bib record/portfolio matching for Kanopy and Swank videos)

  • Follow-up and testing of improvements to Primo OA resources filtering, based on portfolio rather than bib record information - as promised in an upcoming release

  • Investigation in Rialto metadata - where is it harvested from and how often? How reliable is it? How is the data passed to Acquisitions and Alma modules we items are ordered/requested

  • Investigation into Resource and Material type mapping. Figure out differences between the faceted data in Primo and the proposed 'quick link' buttons from a recent release.  Possibly work with IR and DAWG groups to improve mapping in of this metadata field into Primo instances

  • Other issues as desired!

November 30

  • Looking at things like NODEDUPE in the bib records, we have been told that doesn’t work anymore so we need to get the word out to save time and frustration.

  • Christine H/SJSU has been keeping track of issues with e-books and videos and how the title alone is insufficient and will send a summary to Christina H & Jill before their next support meeting with Ex Libris (she did, it was very informative but we did not get to those issues in the December meeting, we are having a longer meeting in January so we can cover those)

  • A webinar is coming up on Dec 11 that interested folks should attend on OA/Paywall (link recording to notes)

  • FRBR has some complicated issues, let’s look at those settings again

  • We uncovered that there is interest in starting an acquisitions group in the ULMS again, or at least an Acquisitions open forum to discuss issues. Christina H will work on this in the new year. Issues include Rialto (wrong/bad records for physical items from vendors; ebook records are great here. Independent publishers seem to be part of the issue here, Ingram and Midwest have great records, for example. Replacements don’t work well either. At least 50k vendors involved!! But need more. Print purchases on standing order do not have correct call numbers. Book awards winners are all on one POL? Midwest will generate records for you with that one POL. International editions are ‘matching on title’ even though they do not have the same title), also the new PO task list interface, updating acquisitions parts of the ERM wiki.

  • Material type vs. resource type is confusing. Requires changes at IZ/locally and NZ/network. Group can look at these settings and make recommendations.

  • This group can also make recommendations on cataloging records to find them easier/better in Primo

  • In addition to gathering Primo matching cases in Salesforce, be sure to look for Rialto cases as well as it all may be the same issue or causing some of the issues

December 5

  • Figuring out how we talk to vendors in relation to data management

  • Keeping in mind that we DO have control of how our data is stored and displayed

  • We also have great power as the CSU and as part of eCAUG (now we are the VC & Treasurer of the group)

  • Get our long list of Salesforce cases in the CSU together to look for trends, issues, examples. Many of these cases have just fallen to ‘in development’

  • Let’s keep in mind that Clarivate is a data broker

  • Issues that are cross-module: OA, Rialto

  • Looking at secondary resource types again (more groups, surveys), how do these fit in with the Data Excellence initiative

  • Trying to get rid of as much duplication as we can

  • Different results that we get when limiting by facets vs. what we see in searches, also that search filter bar that Ex Libris introduced a few months ago

  • Sacramento is having a ‘search engine crisis’ when clicking ‘expand my results’, more in Fall 2023 than before. They have done a bunch of relevancy testing, testing on resource types.

  • The CSU (a sub-committee) should look at our CDI settings, our scope settings, relevancy rankings, etc. since the move to Primo VE & then Rapido…we used to be somewhat the same but this has changed over time and might be causing some of the issues

  • There was an explanation about ‘CDI settings’ – when you turn on a collection the CZ (e.g. Academic Search) there is a setting like ‘search CDI index’ that might not be on for you

  • One issue we see a lot is there is a known item, a seminal work that we search for in Primo, it should be at the top of the results and it is not

  • It has been impossible to separate physical books from e-books in searches/filters, how can we improve this? E-book resource sharing is on the Rapido roadmap so maybe improvements are coming soon.

  • Course reserves searching is its own issue. Course ID searching seems to be a string and not an assumption of a string of letters and then a string of numbers (e.g. ABCD134), as is usually done in US universities. We should look into how Ex Libris is saving/parsing those.

  • Since there are still no plans to allow us to see each others Salesforce tickets, we should make a recommendation to cc: others on your tickets so we all see the action/updates and can add our own notes. This is currently limited to 5 on each ticket but we can work on this and formalize it for the CSU.

  • Christian W/Sacramento shared an article from U Washington (need citation) about 300 papers in Primo that had some good methodology that we could look at for our own testing (watch recording and check)

  • Someone said short title matching and OA filter fixes are on the roadmap? Or we should get them on the roadmap. (watch recording and check)

  • ULMS Discovery group could have a drop-in session where they could look at site’s setting and improve them for better search