[Archivesspace_Users_Group] Massive, flat collections

Tkacz, Courtney (VMFA) Courtney.Tkacz at vmfa.museum
Sun Jun 25 12:22:19 EDT 2023

Hi Megan -

We recently had similar questions as we migrated our 15,000+ vertical file collection from our old ILS to ArchivesSpace. We eventually landed on importing them all as individual resource records, but we put them into their own repository<https://archives.vmfa.museum/repositories/4/resources?>. There were many factors we considered, which may not be relevant to your collection, but I will share them in case they are helpful.

1. We are about to implement an integration between Alma (our new ILS) and ArchivesSpace, and we weren’t sure if we wanted to import all archival objects into Alma or just resource records. We knew that we needed the vertical files to be visible, so importing them as resource records was the safer option.

2. We have volunteers that barcode our vertical files and would need access to ArchivesSpace to continue that work. Putting all of the records into a separate repository allowed us to restrict volunteer access to that repository only.

3. We had similar problems trying to determine how we would break up the files into different collections if we imported them as archival objects. We debated a lot of options and looked at examples from other ArchivesSpace PUIs but eventually felt that importing them as resource records would be less confusing to our users who were used to seeing them as individual records in our old ILS.

4. We are also about to implement an integration between ArchivesSpace and Aeon, so we needed to consider how the files would be able to be requested. In that case, it actually would have been better to import them as archival objects so we could turn off the option for patrons to place a request at the resource record level. But once we made the decision to import them as resource records, we knew we had to keep that option in place with Aeon.

5. We have lots of vertical files that physically have more than one folder and adding multiple instances to a single resource record<https://archives.vmfa.museum/repositories/4/resources/11698> looked clearer in the PUI. We also sometimes digitize material from the vertical files and felt like it would be easier to find and track that material by attaching digital objects to the resource record (I don't have an example of that just yet).

Once we made the decision to import them as resource records, it was quite a journey to get the information from the ILS into ASpace because you can’t import top containers using MARC (or at least, I couldn’t figure out how to do it). So we ended up using CSVs and transforming those into EAD files. Because we had 15,000+ files to ingest, we contracted with Lyrasis to load the files, which went very smoothly.

We migrated the files about 2 months ago, and haven’t regretted the decision so far, but I realize that’s not a lot of time to truly say that we made the right call.

Good luck!


Courtney Tkacz
Senior Archivist

Virginia Museum of Fine Arts
Margaret R. and Robert M. Freeman Library
200 N. Arthur Ashe Boulevard | Richmond, VA 23220
804.340.1497 | courtney.tkacz at vmfa.museum
Pronouns: she/her

From: Afro Charities Archives Staff <archives at afrocharities.org>
Sent: Thursday, June 22, 2023 4:34 PM
To: archivesspace_users_group at lyralists.lyrasis.org <archivesspace_users_group at lyralists.lyrasis.org>
Subject: [Archivesspace_Users_Group] Massive, flat collections


We're embarking on a project to create a finding aid for a newspaper morgue, which is a vertical file with likely over 150K folders arranged alphabetically by subject.

We know we need to break it up, and the time has come to decide exactly how. The most rational seems to be to use the alphabet. The number of boxes per letter ranges from 1 box (for Q) to 53 boxes (for M). Boxes seem to max out at around 200 folders.

What I'm pondering is whether to create a single resource with separate series for each letter, or whether to create individual resources for each letter. We've gotten advice to avoid having more than 5,000 AO's at the same level to avoid instability in the system, so we may need to break some of the bigger letters down further. We know to avoid using containers and locations as artificial collections.

Before we make these basic structural decisions and I thought I'd throw the question out there. Does anyone have experience working with large, non-hierarchical collections like this? Any advice or examples or lessons learned would be much appreciated.

Megan McShea

IG: @afrocharities<http://instagram.com/afrocharities>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lyralists.lyrasis.org/pipermail/archivesspace_users_group/attachments/20230625/5e28a330/attachment.html>

More information about the Archivesspace_Users_Group mailing list