[Eril-l] Bulk search/Full Text find

Tue Jun 10 15:47:25 PDT 2025

So, having said "I wonder if you could...", I just tried it.  I pasted a formatted bibliography into Copilot chat and asked it to give me an RIS file.  The first attempt was abysmal (entire citation in the title field, and nothing else), but after a bunch of trial and error I ended up with a very workable solution, as long as I had reasonably standard citations.

Here's the prompt I finally ended up with:

"I'm going to paste in a list of bibliographic citations in Vancouver format.  Could you create an RIS file that I could use to import these references into EndNote?

Please be sure to separate the authors, title, publication source, date, volume, issue, page numbers, DOI, URL, and other information into separate fields.  If there is any extra information you can't parse, place it in a note field.  Make sure that the RIS file is valid by including the TY field at the beginning of each record and an ER field at the end of each.  None of the other fields are required: just leave them out if they're missing.

Be sure to notice that the authors are listed by last name and initials, separated by commas.  "McDougall T, Armstrong MB, et al." is three authors who should be included separately in the RIS as "McGougall, T", "Armstrong, MB", and "et al.".

While page numbers are typically numeric, they may have alphabetic prefixes (such as "E123-E234") or be in Roman numerals (ix-xvii).  They may also be missing.  These are not errors.

If there are any citations you can't process, please display them here along with an indication of what went wrong for each.  Do not include any unparsed citations in the RIS file."

Basically, I started with the initial "Could you create ... into EndNote?" and everything else was added in response to something that went wrong.  I've now saved this in my prompt gallery to use again.

For my test data, I started with the bibliography from https://pmc.ncbi.nlm.nih.gov/articles/PMC11412117/ ("Finding full texts in bulk: a comparison of EndNote 20 versus Zotero 6 using the University of York's subscriptions" -- seemed appropriate!), but most of those references are to web pages without authors, unpublished manuscripts, and other atypical things.  So I decided it wasn't a fair test.  But the authors had included (on the publisher site, not in the PMC version) a spreadsheet of articles they had used for their evaluation, so I took a random sample of articles from that, formatted them in Vancouver style (it being the one I'm most familiar with), and had much better luck.

In the end, I could just paste in a batch of citations and it'd give me a downloadable RIS file that imported to Zotero 7 just fine (when the time came, I discovered that I don't have EndNote installed on this machine!).  The records aren't perfect (it's a bit iffy with corporate authors, and what would you as a human do with "et al." in this situation?) but they're definitely as good as I'd do with any manual import or reconstruction, and it's much faster.  I was just pasting citations in from a text file, and was doing batches of 10-20.  If you've got lots, I'd start with small batches and see how it goes.

At one point, it offered to look up DOIs for items that didn't have one.  That worked, but it only did a few at a time and kept stopping to ask if I wanted it to do more.  That wasn't really part of what I was trying to test so I didn't pursue it very far.  It'd definitely be worth looking into to make the full-text lookup more reliable.

Fun!

Thanks to Brock Edmunds at Boston University, whose BLC Forum presentation last week included the example of pasting URLs into a chatbot and getting a bibliography back, and inspired me to try the reverse.

Lynda.

-----------------------------------------
Lynda Howell (she/her/hers)
University of Vermont Libraries
lynda.howell at uvm.edu
https://libraries.uvm.edu/dana

From: Eril-l <eril-l-bounces at lists.eril-l.org> On Behalf Of Electronic Resources in Libraries discussion list via Eril-l
Sent: Tuesday, June 10, 2025 10:14 AM
To: Electronic Resources in Libraries discussion list via Eril-l <eril-l at lists.eril-l.org>
Subject: Re: [Eril-l] Bulk search/Full Text find

My approach would be to get them into EndNote/Zotero, and then use its Find Full Text function.

The trick, of course, is getting them into a citation manager.  Depending on how structured the spreadsheet is, you may be able to import it.   I think EndNote's Find Full Text feature works off the DOI/URL if you have it, so you don't have to get all of the other fields parsed out perfectly.

If the citations aren't importable you could paste batches into Google Scholar, PubMed, or Crossref's search box, get the list of results, and export from there.

I saw a presentation this week that gave an AI chatbot a list of URLs and said make me a bibliography.  I haven't tried it yet, but I wonder if you could do the opposite.

Sounds like a fun project!

Lynda
________________________________
From: Eril-l <eril-l-bounces at lists.eril-l.org<mailto:eril-l-bounces at lists.eril-l.org>> on behalf of Electronic Resources in Libraries discussion list via Eril-l <eril-l at lists.eril-l.org<mailto:eril-l at lists.eril-l.org>>
Sent: Monday, June 9, 2025 7:23 PM
To: eril-l at lists.eril-l.org<mailto:eril-l at lists.eril-l.org> <eril-l at lists.eril-l.org<mailto:eril-l at lists.eril-l.org>>
Subject: [Eril-l] Bulk search/Full Text find

Hi all.,

My colleague has completely stumped me. I have a colleague who is helping a faculty member with a systematic review. She has an excel file of citations (mostly from Ebsco databases)  and is trying to figure out how to get the full text of them without having to search each individually. We have Alma/Primo if anyone knows any tricks for doing it in there or in Ebsco.

Any ideas?

Thanks

Theresa

Theresa Carlson

Head, Content and Delivery Services

Cline Library, Northern Arizona University

928-523-6849

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.eril-l.org/pipermail/eril-l-eril-l.org/attachments/20250610/37dedc1c/attachment.htm>