[slide 1]

The digitization of special collections will become increasingly collaborative as demand for digital delivery rises, resources are stretched thinner, and interoperability improves.  Implemented well, collaborative digitization can demonstrate the relevancy of special collections, enhance primary source research, and foster user communities.

The shift towards digital cooperation is evident across the profession in major undertakings such as the HathiTrust and the World Digital Library.  There is a clear pattern in grant funding: 19 of the 51 IMLS National Leadership Grants awarded in 2009 included the word “collaboration.”

[slide 2]

Though many smaller-scaled programs have long featured partnered back-end and collection development,the RBMS community has recently been urged

[slide 3].

to further involve scholarly users and to implement active engagement strategies in our digital programs.

[slide 4]

The University of Vermont Libraries’ Special Collections is responding to this call by redefining its Center for Digital Initiatives.

The CDI started as the digitization arm of Special Collections.  In 2005, UVM Special Collections applied for and obtained federal IMLS funds through Senators Patrick Leahy and James Jeffords.  Many library units contributed: A systems staff member scanned images, catalogers assisted with authorities, and Special Collections faculty selected and managed projects.  A full-time Digital Initiatives Librarian was hired and over the next two years, she built our entire back end, including a native XML database; designed our site’s front-end; and created an easy-to-use, web-based interface for XML metadata creation.

[slide 5]

The CDI launched in April of 2007 with these collections documenting Vermont history, politics, and agriculture.

[slides 6 - 11: Stop at Toussaint]

By 2008, twelve collections were completed or in-progress, thanks to small-scale collaboration:

[slide 12]

–Staff in the cataloging department create museum-level description for these photos:

[slide 13]

–and fellow in Special Collections created about 140 TEI transcriptions for this collection:

[slide 14]

–This legacy digital project from the 90s, which collocated & transcribed correspondence from multiple repositories, was migrated from SGML

[slide 15]

into the CDI.

–Finally, [slide 16]

the glass lantern slides in our Long Trail photographs collection

[slide 17]

were scanned by and also included in

[slide 18]

UVM Prof. Paul Bierman’s  database of then-and-now images, the Landscape Change Program.

[slide 19]

So in the first round of grant funding, the CDI’s digital library infrastructure was built thanks to a few collaborative relationships.  In the 2008 appeal for a second round of grant funding, however, outreach and collaboration became the CDI’s primary strategy for increased use and effectiveness.  It is now our stated goal that students, faculty, staff, scholars and community members will

[slide 20]

participate as users and creators of unique digital research collections

[slide 21]

in an open, collaborative environment.

With this second IMLS grant, I was hired as the Digital Initiatives Outreach Librarian and was charged with the task of fostering a community that will not only use our resources, but will partner with us to build and refine our content, functionality, and services.  We  are now midway through our two-year effort to create this “open, collaborative environment” and the new mission has become central

[slide 22]

to our Collection Development, Project Management, Site Functionality, Outreach, Publicity, and Assessment.

[slide 23]

Building a community of user-contributors demands collaboration on multiple levels:  within the libraries, across the university, and with other repositories.

[slide 24]

In an effort to help pave the way towards our collaborative digital future, I can will share our vision, strategies to date, and expectations for success.  It must first be stated, however, that collaboration is slow.

[slide 25]

We are just about halfway through the two-year grant; I started in July of 2009.  Most of our planned collaborations are just getting off the ground; few of them will actually be completed within the grant period.  Today is a mere progress report; the true outcomes, best practices, and lessons will not be measurable for a while.

[slide 26]

Consider it our hypothesis that the extra time is worth it.  The expectation written into our grant was that sacrificing the volume and speed of a unilaterally managed production mill will result in an organic and sustainable user community.  That — to reiterate my introduction — collaborative digitization can demonstrate the relevancy of collections, enhance primary source research, and foster stronger user communities.

[slide 27]

The CDI’s path towards a community of user-contributors is first and foremost a collection development strategy.  Shortly after my arrival, I led the CDI collection development committee’s revision of our collection policy.

[slide 28]

We wanted the revised document to be as neutral as possible with respects to disciplines, format, and location.  With the goal of drawing in new constituents, we scrapped references to subjects like “history” or Vermontiana.  Our policy previously spoke only of digitizing traditional archival materials; now it is clear we accept both digitized and born-digital collections in any file format.   We actually deleted the words “Special Collections” in an attempt to include collections “beyond” the physical and organizational entity.

[slide 29]

For example, our Maple Research Collection, which launched in March, includes historic photographs which live in UVM’s Proctor Maple Research Center, not in the library.

[slide 30]

As another example, I am cultivating a professor who took digital and analog photographs of Buddhist ritual practice while on research trips in Asia.  Special Collections may never acquire his physical collection, but why not facilitate access to digital surrogates, if he wants to describe them?

There have been pros and cons to this intentionally imprecise collections policy.  Most users look at current collections, not the policy, so they still assume we are only interested in a narrow range of “old stuff in the library.” We strove for vagueness out of a desire

[slide 31]

to encourage creative new ideas,

[slide 32]

but some potential collaborators (including many library faculty members) wish we more concretely defined desirable content.

So what are we looking for?

[slide 33]

The CDI defined four basic criteria.  We deal, unlike other digital libraries, only with cohesive collections (not individual items or indefinite, ongoing projects) which are broadly-valued, which do not already exist elsewhere, and which are enhanced by inclusion in a searchable and browseable digital library (as opposed to on a faculty site or within a Blackboard course).

[slide 34]

Our new collection policy details our criteria for Value, Use, Added-Value functionality, Rights & Permissions, Preservation concerns, and Technical Feasibility.

[slide 35]

In November, we announced a collection proposal process by which anyone can suggest and help create a new collection.  This allows us to build collections according to the needs of current or potential users, effectively creating the community of user-contributors envisioned in our grant.

[slide 36]

Instituting a proposal process demands both traditional cultivation strategies and innovating branding.

[slide 37]

Some of my most fruitful collaborations started by just showing up.  Librarians’ faculty status at UVM makes it easy to attend events and make good connections.

[slide 38]

We marketed the proposal process — through flyers, campus newsletters and direct appeals — to library liaisons and UVM faculty as way to better serve their academic departments.

[slide 39]

We also recently pitched the proposal process to library faculty as fertile ground for scholarship.  Working on collections can develop subject-area expertise and publishable research.  There is also ample opportunity to contribute to process-oriented library literature.

[slide 40]

We also think collection creation is ideal graduate student work, and are speaking with a new interdisciplinary graduate program at UVM about creating one such a culminating experience option for Master’s candidates.

[slide 41]

We did not anticipate a self-starting undergraduate to respond to our call, but we are working with a Junior to establish an Independent Study in which he will create a CDI collection of historic materials from a local public library unable to support digitization themselves.

[slide 42]

From the start, we envisioned receiving proposals from other institutions on the premise that, as one of the only digital libraries in Vermont, we can match our infrastructure and to others’ valuable content.  We also hope to build a critical mass of materials, and to make interesting links between collections which may be physically separated.

[slide 43]

Of course, many smaller institutions need extra funding even to collaborate with us.  As a result, I have ended up assisting several institutions research and apply for small grants.   For example, the Brooks Memorial Library should hear soon about a Windham County Foundation grant for $8,800.  If they don’t receive it, they won’t be able to pay a part-time staff member to work with us on their Porter Thayer Photographs collection.

[slide 44]

More challenging is the fact that many small historical societies and museums are not professionally staffed, have little control over their collections, and – frankly – see collaborative digitization as, well, an excuse for a free consultant.  We need to gently manage expectations about what sort of support we can provide — otherwise, inquiries into digital collaboration could easily turn into archival surveys or processing projects in their own right.

[slide 45]

Perhaps the biggest concern when partnering with other repositories is branding.  We feel that compared to other digital repositories, we are willing to offer a lot of co-branding, since that demonstrates our collaborative nature. Way to accomplish this include displaying logos within the CDI’s existing display

[slide 46, 47, 48]

or allowing partners to publicize a URL on their domain name which redirects to the CDI.  However, we understand that partners may find the CDI page,

[slide 49]

with its UVM branding, more off-putting

[slide 50]

than a neutral site design like Digital Scriptorium.

Different potential partners will respond differently:

[slide 51]

We have signed this agreement with the Vermont Folklife Center to “determine the appropriate amount of co-branding and customization.”

[slide 52]

The result will likely become our template for collaborative design.  For many of our partners, branding has not been an issue.

[slide 53]

However,  one Vermont college with a very notable photograph collection decided, after meeting with us, that a CDI collaboration would not be mutually beneficial. They prefer to find a digital solution exclusively branded with their identity, which we completely respect.

[slide 54]

After all, we only want to partner with institutions when it will be mutually beneficial.

[slide 55]

This flowchart illustrates our collection development process.  We have done, I think, a good job of systematizing a nebulous undertaking. Each collection idea has its own needs & each partner brings different resources,  so our collection development process is partly a conversation and partly a negotiation.

[slide 56]

I work closely with potential collaborators to submit a collection proposal to our Collection Development Committee.  The proposal form is posted on our website, but we’ve never received a proposal out of nowhere.  Cold contacts usually send an email first.  In fact, as a result of a CDI Collection Development retreat in May, we decided that the current form, with its 12 questions based on our collection development criteria, may be too daunting,

[slide 57]

so we will soon move to a simpler, open-ended intake form which only asks “What is your idea?”  We hope this will encourage more contacts.

[slide 58]

When the Collection Development committee receives a proposal, we go through a two-stage selection & prioritization process.

[slide 59]

The Evaluation Checklist is a series of initial yes/no questions such as “Have rights and permissions been documented or obtained?”

[slide 60]

For a proposal to even be considered, all answers must be “yes.” Just getting proposers to this point can require a lot of guidance and communication.

[slide 61]

The committee’s real work is in prioritization.

[slide 62]

We developed a Prioritization Criteria document based on our collections policy,

[slide 63]

and created a weighted Prioritization Scorecard which measures those criteria.  Value and Use are weighted most heavily, and we give extra consideration to UVM teaching and research.  We also factor in the expected impact on our resources.

Like most library collection policies, the overall prioritization score is handy for two reasons -

[slide 64]

if it comes to the point that we are flooded with proposals, we have a method for allocating resources and planning workflow.  This has not been a problem thus far.

[slide 65]

The scorecard is also useful in case our collection development decisions are challenged.  For instance,

[slide 66]

I am the project manager for the Kake Walk at UVM collection about an 80 year minstrel tradition which last until 1969. This collection is doubtless our most controversial to date, but it actually received the highest prioritizatoin score yet, meaning we have documented why it is valuable and useful to digitize.

[slide 67]

We have received 10 proposals since the process was introduced in November. Five of these have moved into project development.  All of the forthcoming collections have collaborative elements:

[slide 68]

As I’ve already mentioned, we are collaborating with the Vermont Folklife Center to deliver four audio collections.

[slide 69]

VFC already digitized these materials with a GRAMMY Foundation grant, but they are unable to deliver those files effectively.

[slide 70]

During the upcoming Civil War sesquicentennial, UVM will partner with the Vermont Historical Society, the Peacham Historical Society, and other repositories

[slide 71]

to collocate materials documenting Vermont’s significant contributions to the Civil War.

[slide 72]

We will also collocate the library’s comprehensively digitized manuscripts with those held at UVM’s Fleming Museum.

[slide 73]

This collection will be contributed to the Digital Scriptorium, once they move from Columbia to Berkeley.

OAI Harvesting is one reason we can even envision a collaborative digital future.

[slide 74]

Similarly, we will contribute our maple collections to the National Digital Library for Agriculture.

[slide 75]

I mentioned the Kake Walk collection.

[slide 76]

I am currently co-teaching a three-credit online Ethnic Studies course entitled “Curating Kake Walk: Race, Memory, and Representation.”  After studying Kake Walk itself, students will contribute to the digital collection by voting on the thumbnail image & working in groups to apply subject headings and write series-level scope and content notes.

[slide 77]

By collaborating, students will understand the subjectivity of descriptive metadata creation, will produce higher quality work, and will ensure community input on this controversial collection.  This course models turning students into user-contributors.

[slide 78]

When a project moves into project development, a Roles and Responsibilities form is signed.

[slide 79]

In this document, workflow and deadlines are agreed upon for each stage of project development.  It is worth noting that collaborating institutions should be responsible for their own permissions and reference requests.  Although all files must sit on UVM servers, we can provide copies on hard drives for our partners’ reference, instruction, and publicity needs.

[slide 80]

Collaboration requires collaborative tools.

[slide 81]

I have completed several grant applications using Google Docs.  My students use Blackboard and Pirate Pad to communicate.

[slide 82]

In the CDI, we pay $50 / month for Basecamp project management software.

[slide 83]

Its messages, to do lists, time tracking, writeboards, and file sharing are invaluable.

[slide 84]

The very first collection to go through the proposal process, Medieval and Renaissance Manuscripts, just launched yesterday.

[slide 85]

When a collection is finished, we launch it with a targeted outreach and publicity plan.

Events and video projects lend themselves especially well to fruitful, collaborative publicity.

[slide 86]

We have partnered with three different local television stations to show our PSAs, appear on talk shows, and create program content.

[slide 87]

Public events are especially key to connecting with users since digital products are by nature decentralized and often experienced individually and remotely.

[slide 88]

From a wildly popular Maple Cook Off

[slide 89]

to a film screening of Spike Lee’s Bamboozled, events are integral to our vision of a community of user-contributors.

[slide 90]

But how will we know when we have a vibrant, organic, sustainable community?  How can we know we have met our goal?

[slide 91]

We will continue to use Google Analytics to see what gets used and how often.  We can tell where users are located and, to some extent, how they found us.

[slide 92]

We see increased traffic as one sign of success.  We can measurably see, for instance, traffic peaks near specific outreach events.

[slide 93]

We are also gathering data to see if target audiences are aware of or use our resources and services.

[slide 94]

From this year’s survey, we can confidently say – not that many people knew about the CDI.

Of course, the survey itself is an outreach tool,

[slide 95]

and responses about anticipated use in the coming year are much more encouraging.

[slide 96]

We are collecting this data with the expectation of seeing more recognition over the next few years.

[slide 97]

One great outcome of the survey is that 124 students and 112 faculty members voluntarily shared contact information with us.  We will use these names to compose focus groups this fall for improving site functionality such as a “bookbag” of saved records or Zoomify magnification of images.

[slide 98]

However, this OCLC publication calls for more than just quantitative metrics.  We want to see more visitors and to become a more recognizable brand, but we also want to know that our users are really creating the kind of “open, collaborative” community envisioned in our grant.

[slide 99]

For this reason, feedback from collaborators and partner institutions, course evaluations, student work on CDI-related assignments, emails to the CDI, comments on the CDI site, media coverage, and programming success may more fully describe whether or not we are meeting our goals

[slide 100]

of creating a collaborative, digital community.