ARLIS/NA 47th Annual Conference

(NOTE: Attendees should come prepared with an internet-connecting laptop and a sample dataset, if possible. Two sample datasets will be provided, if attendees cannot bring their own.)

Collections as data has quickly become a buzz phrase over the last several years as more and more libraries begin to identify collections that translate well to digital humanities projects. In July 2018, the Mellon Foundation awarded a $750,000 grant to Thomas Padilla and the IMLS effort towards “developing, describing, providing access to, and encouraging reuse of collections that support computationally-driven research and teaching.” This grant comes on the heels of work Padilla and the IMLS have already accomplished on an idea that digital collections can be more than simply surrogates of their represented physical items or born-digital items and that the datasets should be offered as machine-actionable data ready for computational research methods. Padilla calls this initiative, “Always Already Computational,” and it was formalized by a series of forums, conversations and workshops into the Santa Barbara Statement on Collections as Data (2017).

Many of the DH projects to adhere to the Statement have been in fields outside of art history and libraries—American history, journalism, public history, and social science. However, collections as data principles are inherently valuable for art libraries interested in leveraging their own collections to support computationally-driven research and engage new users. What can this initiative mean for you and your library collections? Presentations detailing two institutions' work towards collections as data will give first-hand experience to the challenges and strategies of utilizing linked open data for new, exciting datasets. Presenters from the University of Southern California will discuss a first attempt at creating a larger dataset that merged different collections in the context of the LA Arts Datathon. The LA Arts Datathon was a one day event that featured different tracks and activities—all with the aim of engaging the public to think critically about how art and data intersect. How to scope a project, normalize the data, and lessons learned will be some of the insights from this presentation. Presenters from the Autry Museum of the American West will continue this theme, discussing a case study exploring the development and future of the Museum’s ongoing project linking art and images with the publications in which their reproductions appear. This implementation of linked data enables new appreciations of how collection items are used in current research and how current research is interpreting art, material culture, and historical images. Looking to the future, they will discuss how linked data improvements may enhance context, discovery, interpretation, and research.

More than half of this panel is entirely interactive, with worksheets from both presenters geared towards attendees’ own collections. Participants will develop an action plan that includes ideas for potential collections, identifying useful tools for this process, and strategies for thinking about collections as data and building a community of practice. Instruction from the presenters will include a walk-through on the basics of metadata, including how to standardize and prepare it for teaching. Additionally, participants will work with presenters to develop a workflow for creating a collections as data project, including defining tasks, roles and responsibilities, and addressing sustainability concerns to ensure the project is supported going forward. Discussions will address participants’ questions and concerns as they analyze their data and their collections.

Learning Objective
Participants will have an action plan for exploring and engaging with a collection as data within their own institutions.