Over the past several years, the Digital Scholarship Group has developed and collaborated on multiple open access digital scholarship (DS) projects. Each DS project is developed with specific goals in mind and is built with different content, frameworks, programming languages, and design elements.
In order to make our work discoverable and accessible, we promote and disseminate our projects through various means, including on our library and digital scholarship websites, social media, professional organization mailing lists, conference presentations, and literature, as well as share our code in repositories and with discipline-specific communities and projects (such as the CANTUS Manuscript Database).
One of the ongoing challenges of creating DS projects has to do with preservation and sustainability. While many of our projects are developed from the bottom up (Burns Antiphoner) and require significant programming or application development, others make use of existing infrastructure (John La Farge, Stained Glass in New England) and can be edited through a content management system. Our projects all currently live online and are being maintained, however we want to make sure that the core components remain accessible, preserved, and can be reused for years to come. This is why we are engaging in multiple approaches to document our projects, store our code, (meta)data, and other content types. We have a BCDigSchol GitHub repository, which we use during a project’s development to keep track of our code and make our work accessible and reusable. Although digital objects featured in our projects, such as Cristóbal de Morales’s Missarum liber primus (Lyon: Moderne, 1546) in the Morales Mass Book or audio files of traditional Irish tunes in The Séamus Connolly Collection of Irish Music are accessible in compressed format on the project sites, any raw image or audio files that require preservation are stored on a preservation quality server rather than in the GitHub repository. Our goal is not to store and preserve each and every piece of a DS project in a single repository, because there are already mechanisms in place for preserving digitized images and audio, but rather to identify which data and content types of a DS project would most benefit from a shared common space.
To that end, we recently began exploring the Open Science Framework (OSF) as a repository where we could share data and content types from our DS projects that would be most meaningful to other scholars and practitioners. We looked at the OSF because of the potential it has as a repository, a scholarly commons, and an open data set. OSF’s potential as an open data set is a promising endeavor and is being developed in partnership with ARL’s SHARE initiative. If the SHARE initiative is successful in meeting its goals—identified as “building a free, open, data set about research and scholarly activities across their life cycle”—then one of the benefits for DS projects with content in OSF is that the projects will be automatically aggregated and made widely accessible and discoverable through a single search interface.1
We selected The Séamus Connolly Collection of Irish Music as a use case for OSF, because this project had a number of content types and documentation that could be brought together into a shared space.
We linked the project’s GitHub repository to OSF so that all of the Omeka files can be easily accessed. In addition, we used a Python script to export all of the metadata from Omeka to CSV files and added them to the repository. Each song and tune transcription, which is displayed as a PDF on each item page in Omeka (for example, see For Betty Killoran) was converted to MusicXML and the files were uploaded into the OSF project.2
Although the work on The Séamus Connolly Collection of Irish Music itself is complete, we are currently pursuing exploratory work related to the collection and focused on developing linked open data and network visualizations. Within the main Connolly OSF project we were able to create a component (module) for the Linked Data Exploratory Project where all the datasets and documentation were uploaded and made public.3
In addition to the potentials already highlighted, here are some additional features of OSF, which appealed to us:
Identifiers – OSF enables you to easily generate a DOI and ARK for your entire project. Although each DS project has a domain name and stable URL, providing a DOI will further ensure that a project and components (hosted in or pulled into OSF) are discoverable, identifiable and citable.
Attribution – DS projects are often collaborative endeavors developed by many individuals from different backgrounds, departments, and roles. This feature enables you to identify all of your collaborators who contributed to the project. You can specify those who are bibliographic or non-bibliographic, as well as give them different credentials within the project repository. While we identify collaborators and roles for DS projects at Boston College Libraries on their respective project sites, one thing that is missing in OSF is the ability to identify or assign specific roles to the collaborators. It could be as simple as a tag, which administrators could create and assign in order to identify roles, such as project manager, developer, metadata librarian, etc.
Sync with external services – Another aspect that drew us to OSF is that it is interoperable with external services so that you can easily sync files from your code, content, and citation management repositories, such as GitHub, Box, or Zotero. These files will then appear in your OSF project repository.
Analytics – Basic analytics enable you to view number of visits, forks, external links to your project, template copies, top referrers, and popular pages.
Some of the other features worth noting are the Activity Log where all activities within the project repository are visible making it easier for a project team and anyone visiting the site to see how active the repository is and who has made what type of changes or contributions, as well as a Wiki space and Tags that can be added to increase greater visibility to your project site.
We plan to continue to explore the potential of OSF and upload additional data and contents for select digital scholarship projects in order to make them discoverable and accessible.
1. More details about ARL SHARE can be found online.
2. We would like to acknowledge Michael Scott Cuthbert who encouraged us to output the transcriptions in MusicXML and ran the initial conversion.
3. A current interactive version of the Connolly Collection network visualization can be explored online.