A while ago I tried to capture most of my general recommendations for AtoM migrations in the following slide deck:
A couple of the suggestions in those slides may be helpful to your colleagues - specifically these two tips:
- Importing the sample AtoM CSV as a crosswalk to better understand how AtoM's CSV template works, so you can make a proper crosswalk to the DBTexworks exports, as shown in slides 10-11
- Looking into the use of OpenRefine as a tool to handle data wrangling, as shown in slides 27-29. It can be really helpful for splitting or combining CSV columns, and there are a TON of free resources online - usually, you can just search for whatever you are trying to do and find a ready-made example you can adapt.
It's been a while since I have led an AtoM migration project, but here are a couple other general suggestions not found in the slides:
One of the biggest challenges we see when migrating from older systems is that these rarely have separate records for authorities. This can sometimes mean that if the same creator appears in more than one fonds, each has a different biographical / administrative history statement that is attached directly to the fonds, rather than being normalized and shared in a separate record. If you import these as-is, AtoM will end up making duplicate authority records (i.e. 2 authority records, each with the same authorized form of name, but with different histories). OpenRefine (or even just a good search in a spreadsheet application) can help you find such occurrences - the archivists will need to decide what to keep as the official history going forward, and make sure that any information in the existing history statements that is actually about the records / fonds / etc and not general information about the actor is moved into the proper description instead, so it is not lost.
In some cases, this is the same for access points such as subjects and places - if these DB/Textworks installations do not have dedicated management of access points, then there may be near-duplicates to resolve in the data review and cleanup phase - for example, one fonds has the term "cars" as a subject, while another has "automobiles" and yet another "vehicles," etc. Once again, OpenRefine or even just a good spreadsheet application can help you find these instances, and the archivists overseeing the migration will need to decide how they would like to normalize the terms before proceeding, or else commit to reviewing and addressing near duplicate terms in AtoM post-migration.
Good luck! And remember, if any of these institutions would like paid data migration support, Artefactual offers such services:
Cheers,