Building Sustainable DH Projects w/ Re-usable Components

How have DH projects evolved from “boutique” to modular/sustainable over the years? What are faculty, librarians, and academic technologists at your institution or organization doing to create sustainable, adaptable models that can be used by multiple DH projects? Why have some of the DH projects (Bamboo, etc.) not gained more momentum? How do we share common frameworks and components to make DH projects simpler to build and maintain? How do we resource projects that require maintenance and enhancement over time, and how do we reduce costs by leveraging existing tools and standards?

Categories: Coding, Collaboration, Funding, Project Management, Session Proposals |

3 Responses to Building Sustainable DH Projects w/ Re-usable Components

  1. I really like this topic, and would be interested in expanding it (if warranted) to include a discussion of cooperative approaches to providing support for DH research. Is anything like this happening? What opportunities and potential pitfalls are there if multiple institutions agreed to support one another’s faculty and staff in specialized services and skill areas (i.e. metadata standards, database design, text mining and visualization, topic modeling, GIS data management and analysis, 3D architectural reconstruction, content management system selection/optimization/hosting, peer review and publication, etc.) ? In other words, is forming networks of cooperating specialists across libraries one path toward sustainability? I am not currently working as a librarian so only have an outsider’s viewpoint but would be really interested in what others have to say about it, in case there might be ways I could help provide information and encouragement to librarians interested in pursuing such a strategy.

    If this seems more appropriate to consider as a separate proposal, that’s okay with me, too, but I woke up thinking about this and your post struck me as closely related.

  2. I like this topic as well. In my application for THATcamp I mentioned I was generally interested in ways of making it easier for digital humanities projects to “transition” between active scholar-driven initiative and library-maintained resource, and vice versa. (Lasting DH projects can go through multiple stages and stewards, after all.)

    Putting on my computer-science hat, I suspect that a big factor in long-term sustainability is robust, well-structured and easily shared data. If you can encode the knowledge embedded in the project in portable forms, then it can move between components and infrastructures as the knowledge remains valuable, and as new kinds of useful components and infrastructures arise. So I wouldn’t mind seeing some discussion specifically focusing on sustainable data in digital humanities projects, though I don’t know offhand whether it fits best under the umbrella of this proposal or is best discussed on its own.

  3. Dot Porter, with whom I work at IU, came away from last year’s DLF with a couple of great rubrics from Emory and NYU for cultivating, managing and sustaining (or not) DH projects, which she and others at IU have since morphed into “levels of services” model that integrate in a building block approach production-level “digital library” workflows and services with specialized digital research needs of scholars. The “levels” serve as a framework for the back-end (i.e., technical) and the front (i.e., consultation), but we have yet to implement this model in practice. Here’s a quick summary (you may notice that we have adopted NYU’s approach for level 3):

    * Level Zero, Commodity Tools: Use of open source and site-licensed software tools supported by the Libraries and/or UITS to provide personal digital asset management/digital collection building, text analysis, data mining, statistical analysis, reference management, and other functions useful to digital humanities and scholarly projects.

    * Level One, Fundamental Services: Use of standard digital collection and repository services provided by the library and DLP with no minimal customization, including support and training.

    Level Two, Customized Services: Building on fundamental library services, customizations should be designed and developed in such a way that they contribute to the ongoing development of our fundamental services and technical infrastructure.

    Level Three, New Developments: Development of new software to support scholarship. New developments should be designed with the aim to be “first of a kind” rather than “one of a kind,” and could lead to the development of entirely new services and expanded technical infrastructure.

    I am in principle a supporter of this model, but I think this model and others like this one may be fundamentally flawed in that our library at least is still creating barriers to digital scholarship. We would need to get through at least two levels before we can truly engage the scholar, not to mention the problem with assuming that the underpinnings of the first two levels are indeed in place.

    I see threads of all three posts before this one manifesting here, but I wonder if some part of the libraries, in collaboration with other groups, centers, people across campus, can get to the point in which it is perfectly valid to invest resources in a specialized digital research project regardless of broader impact or long-term sustainability. Perhaps more weight should be given to transormative scholarship that is initiated by the research (research needs before content) versus by the libraries (content, then, maybe, research needs).

Comments are closed.