Update on Preservation metadata task force
(PREMIS element set & usage guidelines)
The task force on a metadata element set for preservation has met and begun its work. This is just to give you all a quick update on what we are doing and planning to do.
Our deliverables are:
1. A PREMIS-based metadata element set and usage guidelines for use in preservation of digital assets at Yale
2. A report of the investigation: its methods, commentary and recommendations
Our meeting arrangement:
1. Consult with the task force on a metadata element set for preservation advisors via monthly meetings and individual consultations as possible or needed. Next meeting of whole group set for July 10.
2. Meet with small working group weekly. The first meeting of this group was May 31st.
Our process plan:
To create the element set and guidelines, we are working on a top-down analysis of the PREMIS model and its semantic units in conjunction with several bottom-up use cases. The aim of this two-pronged approach is to gain a logical understanding of the PREMIS model, a practical understanding of our needs, and concrete experience with implementation constraints and opportunities.
To understand the model from the top-down, each member of the small working group is reading the PREMIS model documentation, related articles, and discussing them. From this, we'll develop a draft of the most likely-to-be-relevant PREMIS semantic units (elements) and begin to test the draft with others. To gain experience with our needs and concrete use of the PREMIS model for digital materials in our collections, we have identified existing digital collection projects or programs at Yale as possible use case partners. One for data sets, one for texts, and two for still images. With each of these projects we expect to use a Fedora test bed maintained by David Gewirtz, one of the task force advisors, as our repository. This Fedora repository serves us as a logical proxy for a Yale preservation repository.
The existing digital collection projects or programs are:
1. Data sets:
Social Science Data Archive's migration from its current environment to Fedora. Our collaborators are Michael Appleby, ITS, and Gretchen Gano, SSLIS. We are working with a small sample of social science data sets that use Data Documentation Initiative (DDI) metadata for the studies and derived Dublin Core (DC) descriptive metadata. This use case partnership is already established and we have one instance of a study in the Fedora test bed using PREMIS.
2. Texts:
YUL Preservation Dept. Reformatting and Media Preservation's digital texts. Our lead collaborator is David Walls ( a member of the task force's small working group). We'll be working with a small sample of digital texts that result from the reformatting process. Initial texts will be PDF files. Descriptive metadata for the texts is available through the MARC format records for the print originals. This use case partnership is established, and sample set of PDF files is being selected.
3. Images:
A. Yale University Art Gallery's digital still images from its collection management system (TMS). Chief collaborator would be Thomas Raich, YUAG, and a task force advisor. We'd like to be working with a small sample of digital still images from the Gallery's collection. Descriptive metadata for the images will come from the TMS system. This use case partnership is not yet established. We have contacted Thomas, but not heard back yet.
B. DPIP/Arts Library's Classics Dept. project's digital still images. Chief collaborator would be Karen Reardon. We'd be working with a small sample of digital still images from the project to digitized images from collections of the Classics Dept. Descriptive metadata for the images would come from the DPIP/Arts Library work on the images as they are migrated to the Insight system. This use case partnership has not been established, but we have been working out details with Karen Reardon.
These use cases will build our understanding of our needs for a variety of content types, format types, and use environments--ITS, the Gallery, the Library. We'll relate that understanding of our needs to the PREMIS model and use the experience we gain to modify our draft of the relevant PREMIS elements and develop our usage guidelines.
We also want to consult with other PREMIS implementers internationally via the PREMIS Implementers' Group (PIG). There is a discussion list, a "play pen" or development area, and a registry of implementation projects. I'm on the discussion list; we have contacted Rebecca Guenther, LC, (LC is the PREMIS maintenance agency) about registering our PREMIS work (and thus gain access to the development area). However, we are not sure that our work constitutes even the start of an _implementation_ of PREMIS. Although we have adopted a working approach that relies on actual use cases, we are perhaps best seen as a prelude to an implementation of PREMIS at Yale. We think the group and the empowering committees and IAC may want to think about this and make a decision before we register with PIG.
The working group also has discussed our interest in bringing a PREMIS expert or experts to Yale to develop our understanding of the model. Rebecca Guenther and Priscilla Caplan, Florida Center for Library Automation, (Rebecca and Priscilla are the PREMIS co-chairs) are doing a 2-day workshop in Glasgow this July (sponsored by Digital Curation Centre), preparing training materials, and looking for venues in the US (Washington, DC, etc.) to conduct the workshops. Would Yale be interested in hosting? There are a number of angles to be considered and discussed--when, who would attend (how many of us), where, and funding. The tutorial sessions planned for Glasgow include:
Introduction to the PREMIS Object Model
Walk through PREMIS Objects, Events, Rights and Agents
Implementation issues
XML Schemas for representing Data Dictionary elements
Real-life examples of implementation from UK institutions
The next full task force meeting (with advisors) is scheduled for July 10th. We'll provide another update before that meeting (late June) and then another just after the meeting with advisors.