During FY15, we began to develop workflows and procedures for creating disk images of carriers and for removing files from these carriers. A number of collections were completely imaged, and the files are available for processing.
As each collection is worked with, we'll revisit and revise our currently basic workflows for processing born digital files.
What follows is a DRAFT workflow for working with files removed from carriers.
This is based on Susan Earle's July 2015 work with the Records of the HRPBA Oral History project, and on Cat Holbrook's work with the Mary Bunting Institute records in fall 2015. Every collection is, as we know, DIFFERENT (!) and workflows for different kinds of records and situations will vary. See below for how this workflow worked with different models of collections.
SURVEY
- Digital archivist Jen Weintraub is in charge of creating disk images/removing data from carriers. She should have completed this work by the time the collection is ready to be processed. Check in with her about where the working files are kept (on the R drive).
- During your survey process, view (as best you can) the born digital files to be processed. Schedule a meeting with Jen and possibly Pablo Morales-Henry to assess the born digital files. They should be able to provide archivists with a list of files that were removed from carriers; and a list of carriers that could not be read. If carriers could not be read on the first try, archivists should assess (in collaboration with Jen and Pablo) whether or not we should pursue other options for reading those files.
- You and Jen may start by looking at the files in FTK (Forensic ToolKit software), especially if it is a big collection. You will be able to mark the ones you definitely want to keep and process, the ones you definitely do not want, and then ones you want or need to look at in more detail and assess further. In the future, we will also be able to note which files have restrictions and what kinds of restrictions using bookmarks in FTK.
- Once you determine which files you want to keep and look at further Jen will export them to the Digital Collections Storage on the R drive. Some files can be viewed using Quickview at your own desk, and some can only be viewed by using FTK Imager on a forensic workstation.
APPRAISE
- Determine what kinds of files these are. Are they closely related to the paper files? Are they transcripts, drafts, diary entries? Do they have an intrinsic arrangement? Do they have folder titles that make sense? If you believe these files are possibly exactly the same as paper records in the collection, is there value to keeping them in electronic format? Weigh the possibility of time-consuming matching against access and restriction issues. For example, if the born-digital file will be restricted for 50 years, it may be worth your time to check to see if it is in fact a duplicate, since keeping dark digital files may be complex.
- Decide which files to keep and which to discard.
- For files you don’t want: for collections for which we look at in FTK, you may just ignore them and not export them, keeping them in the disk image. For files you look at and then decide you don’t want, you can simply delete them. We can always retrieve them from the disk image.
PROCESS
- Describe electronic files in the finding aid. If they are related to paper files, intermingle them. If it seems more appropriate to have an entire series of electronic records, that's fine too.
Some things to consider when determining how to describe the files:
Are the files restricted in any way?
Can you provide access to them directly out of the finding aid?
Are they complementary to paper files (e.g. are they drafts of a book project that already has a subseries or a number of paper files listed)
Are they arranged in meaningful folders on their carriers? Or are they just put on the carrier as a means of transport?
2. Access considerations. Can these files be linked through the finding aid? (Are there restrictions, has the donor agreed that material can be made available online?)
If yes, determine (with Jen, Pablo, team lead) if there are normalization needs, what are the available ways we can deposit into DRS. E.G. If you have word processing files and plan to create individual links from the finding aid, convert the files to PDF/As, rename individual files with the corresponding E# from the finding aid, and deposit files into the DRS, and link from finding aid. Consult with team lead, Cat Holbrook, and Paula Aloisio on these steps as needed. [Also information to come: how to normalize and provide access to non word processing files such as images etc.]
If no, work with Jen to correctly label, restrict, and store files. Work on how to provide in-house access to come in 2016.
MODEL A: Born digital files are closely related to the paper files, and (due to their file uniformity and small number) can easily be transformed into PDF/As, deposited into the DRS, and linked out through the finding aid.
EXAMPLE: HRPBA Oral History project: http://nrs.harvard.edu/urn-3:RAD.SCHL:sch01503
These born digital files were all word processing documents, and Susan knew from the disk names and the collection materials that they were intimately related to the other records (primarily audiotapes) in the collection. She reviewed the files on R, and determined that each electronic transcript would be a separate "E folder" in the finding aid, listed after the audiotape of the interview. She (with help from Cat?) transformed the word processing files into PDF/A files, Cat deposited those files in the DRS, and links to each file were added into the E folders in the finding aid.
MODEL B: Born digital files from the Mary Bunting Institute Records
Bunting Institute records are almost all under some restricted access. Linking files directly from the finding aid is only available in a few series.
Cat located many disks during processing, some of these disks were unreadable on the first try, and some were identified (after a brief file review) as being not from the Bunting Institute.