Reflection and Selection: Creating a digital project archive

At the end of any project, the time comes to wrap up work and, hopefully, to prepare project outputs for dissemination and archiving. We’re working with the folks from CTREP to create the digital archive for the Learning Landscape Project. It’s an interesting process – while working to pull together the materials for inclusion, I’m also thinking about what we learned on the way, mentally summing up the project’s achievements, as well as the things we could have done better. I’ve reflected that this process of archiving is, in many ways, the process of creating the institutional representation of our project: its official history. I’m not suggesting this is the “best” story, or the “only” story, to tell about our work – there are, of course, many stories to tell. But this collection will constitute (we hope!) some of the most accessible documentation of our work, so we want it to be diverse in content, and we want it to be as thorough -- as rich -- as possible.

CTREP is a project under JISC’s Repositories and Preservation programme. At Cambridge, the CTREP team is working to integrate our VLE, Sakai/CamTools, and our institutional repository, DSpace@Cambridge. (Up in the wilds of Scotland, our CTREP colleagues at the wonderfully-named University of the Highlands and Islands are doing related integration work with TETRA / Fedora). The idea is that DSpace will appear as “just another folder” in the Resources area of a CamTools site, and that pushing items from Resources into Dspace will, eventually, be a simple matter of drag-and-drop. Metadata will be pulled in automatically, along with each individual resource item—although I’ve got a lot to learn about how that works, exactly. Today, I’m working on creating a table that maps the metadata element set used by our project (which was based on Dublin Core) to the Resources “Edit Details” fields in Sakai/CamTools. Much of this is straightforward, though some Sakai-tastic idiosyncrasies have crept in.

Of course, there is a lot of workflow that has to (or: should) happen before a digital resource held in Sakai/CamTools is ready to go into DSpace. In part, this is due to complex, organic human factors: it sits at the very core of what happens when people work together over time. The LLP team used Sakai/CamTools as the home for our “working” project site, and as the worksite grew, along with the research activities, the initial structure we’d designed for it became outmoded. New tools and folders were added. People in the team took ownership of different parts of the site, and redesigned them accordingly. As we added people to the site, some of whom had only a peripheral interest in the project, it became more important to create “private” areas to which only the core team had access. Taken together, all this conspired to increase the complexity (and un-usability) of the site. So we established a second, parallel site to serve as the internal institutional project archive site. A clean slate, if you will. It’s from this second, archival site that items will eventually travel through into DSpace.

Various editorial processes happen before an item even makes it through from the project’s worksite to the second, archival site.  Inevitably, our archival site does not reflect the messy “reality” embodied in the project worksite, and perhaps especially in the project Wiki. But an archive has to balance the desire to keep everything with the ultimate goal of usability (which I take to be a higher-order goal, encompassing the sub-goal of “discoverability”). It’s a process of reflection and selection that strikes me as being “portfolio-like”, even though a research project archive is rarely represented as a “portfolio” of work.