Presentation 1
More than Metadata: Where the Rubber Meets the Road in DLS Migration
So you've gone through the long and arduous process of selecting a new digital library system, implementing the technology, and now it sits waiting to receive your collections. Where do you start?
This working session covers several use cases that detail the complexity of large-scale digital library system (DLS) migration from the perspective of three university libraries and a statewide academic library services consortium. Each will describe the methodologies developed at the beginning of their migration process, the unique challenges that arose along the way, how issues were managed, and the outcomes of their work.
Florida Atlantic University, Florida International University, and the University of Central Florida are members of the state's academic library services consortium, the Florida Virtual Campus (FLVC). In 2011, the Digital Services Committee members began exploring alternatives to DigiTool, their shared FLVC hosted DLS. After completing a review of functional requirements and existing systems, the universities and FLVC began the implementation process of their chosen platforms. Migrations began in 2013 with limited sets of materials. As functionalities were enhanced to support additional categories of materials from the legacy system, migration paths were created for the remaining materials.
Some of the challenges experienced with the institutional and statewide collaborative legacy collections were due to gradual changes in standards, technology, policies, and personnel. This was manifested in the quality of original digital files and metadata, as well as collection and record structures.
Additionally, the complexities involved with multiple institutions collaborating and compromising throughout the migration process, as well as the move from a consortial support structure with a vendor solution to open source systems (both locally and consortially supported), presented their own sets of unique challenges.
Following the presentation, the speakers will discuss commonalities in their migration experience, including learning opportunities for future migrations.
Presenters: Jamie Rogers (Florida International University), Lee Dotson (University of Central Florida), Lydia Motyka (Florida Virtual Campus (FLVC), Joanne Parandjuk (Florida Atlantic University)
Co-author: Melissa VandeBurgt (Florida Gulf Coast University)
Presentation 2
Efficient, Scalable Integration of EML Metadata within a DSpace Repository Institutional repositories (IR) provide an archival function which is complementary to the typically more dynamic architectures of discipline specific data portals. In practice, the capabilities and archival workflows within IRs tend toward a flattening of domain specific ontologies into more general purpose data and metadata models, but as libraries endeavor to scale up research data management and preservation services the need to "walk the walk" - to demonstrate a dynamic and robust re-use of data and metadata across disciplines - becomes a central requirement of outreach, instruction, and curation activities. Considered together with the potential for increased IR use as a low barrier means of fulfilling open and public access requirements, it is equally important that processes for IR enhancement consist of minimal customization and novel implementations of existing repository features.
As a case study and to support the preservation of data previously published by the Sevilleta Long Term Ecological Research station, the University of New Mexico Libraries have developed a process for capturing the value additions provided by the host system, the LTER Network Data Portal (
https://portal.lternet.edu/) and batch harvesting the data for ingest into LoboVault (
http://repository.unm.edu/), a DSpace instance and the designated institutional repository of the University of New Mexico. To enrich the archival record, the Ecological Metadata Language (EML) metadata served through the LTER network is mined to reproduce such value added portal features as item level maps, provenance metadata snippets, and pre-formatted citations. Requiring only minimal customizations to the default DSpace configuration, the new workflow capitalizes on the system's batch ingest and templating capabilities to promote enhanced discovery, indexing, and potential re-use via a fuller representation of the source EML.
Presenter: Jon Wheeler (University of New Mexico)