Awardee: Yvonne Elet
Semester of Award: Fall 2018
Materials Awarded: I received DevonThink Pro Office and Devonagent Pro software. I also received a 4 TB RAID 1 redundant database storage device to support the application.
Project Description:
Rationale:
My objective was to implement a document/knowledge management system to manage multiple interrelated book and article projects, new course development and ongoing course upgrades, and the flood of new bibliography relevant to both research and teaching. At this stage of my career, in a sub-field that is extremely data intensive, for which I collect and analyze information in many media, and continue to research interrelated projects that span many decades, the volume of information is overwhelming. Keeping track of what I have, where it is, and making connections among information in different media has become increasingly difficult and time consuming. It is easy to lose track of details and entire chunks of research. I chose DevonThink, an information management system, after also considering Mendeley, Zotero, Evernote, and Citavi. I found that DevonThink is the only one that combines bibliographic support, information/document management, relational database support, and additionally applies AI to make connections among information in different media types. It also encompasses using Abbyy OCR to bring in lots of things as text, which can then be intelligently searched and linked.
Experience:
Implementing DevonThink has been a bigger project than I had anticipated, for many reasons; and therefore I have only partially accomplished the implementation, and am continuing it as on ongoing project. First, DevonThink is a MAC only product, and I am a lifelong PC user (and former IBM systems engineer), so I have had to switch from PC to MAC in order to use it – an ongoing migration. Second, the software itself is exceptionally flexible and robust, on the positive side, which means it is also very complex, with a steep learning curve, on the other hand. This is actually more than a learning curve; it involves making creative decisions about how to map one’s resources, which I’ve found to be an iterative process – the more I understand how the package works, the more I change or tinker with the way I set things up. Third, I already used EndNote for bibliographic support, and EndNote has introduced a new version with major changes that required work on my part – another task that is part of the overall project. Finally, the scope of the project is huge, as I eventually hope to link my documents, endnote libraries of bibliographic references, my database of images, PowerPoint files filled with images for teaching and research presentation, email, and html files. The amount of material and the complexity of the project is challenging; I could use a sabbatical without pressing publication deadlines to focus on this kind of infrastructure task, which is difficult to do while producing content at the same time. (I have had an unusual crush of writing and presentation deadlines over the past couple years, in part driven by the 2020/21 Raphael quincentenary.) In sum, implementation is complex; I am treating it as a long-term project, and breaking it down into smaller components.
Assessment:
This is a complex product, which demands a level of technical understanding, and an investment of time and concerted thought. The user needs to gain a fairly deep technical knowledge of the software’s internal workings (file structures, terminology, and classification methods) in order to make good use of the system. It is much like creating a database, in that the database structure can make subsequent additions either easy or hard, and may require some re-work in order to address an issue (or fix a problem) that was made in an earlier step. Also, one’s research workflow is complicated by the fact that documents and metadata are often replicated across physical storage devices and different applications and operating systems. It seems that one has to conform workflow to the tool, rather than vice versa. That said, I have found the AI connections to be impressive in terms of document classifications and automating “groupings”.
Impact:
It is too soon to say what the impact on my own work will be; it takes an investment of time and creative energy well before one sees the results. I have used this experience to talk with students who are bound for Ph.D. programs about tools they may want to acquire for knowledge and resource management, so they can grow their research, bibliography and resources organically from the start – the ideal scenario.