Managing Digital Assets from Curation to Exhibition

As the Libraries at Texas A&M University continue to accumulate digital assets and cultivate a Digital Asset Management Ecosystem (DAME), an urgent need is developing for metadata annotation workflows and the marshalling of documents from digitization to exhibition. Last year the Libraries presented an alpha version of MAGPIE (Metadata Assignment GUI Providing Ingest and Export) and demonstrated the curation of a scanned dissertation and its export to a DSpace repository. Since then we have built upon MAGPIE to integrate it into the DAME more broadly. MAGPIE is positioned to serve multiple projects including scanned legacy dissertations, historic and modern agricultural serials, and special image collections destined for Spotlight exhibits.

The original use case of MAGPIE was to shepherd previously cataloged historic dissertations from scanning and OCR through additional curation and finally into the OAKTrust institutional repository (IR). However, as additional digitization and curation projects have manifested, MAGPIE has been a natural fit to accommodate the varying types of IR, metadata authorities, suggestion providers, and exporters. The application includes a repository interface enabling publishing metadata and assets into an IR as a single item or batch. Current implementations include Fedora and DSpace, both via their REST APIs. The interface for consuming authoritative metadata has been implemented for Voyager (again, over a REST API) and CSV spreadsheets on disk. The application also has an implementation interfacing with a metadata suggestion service using the National Agriculture Library Thesaurus. Implementations of the exporter interface provide CSV metadata spreadsheets for download and Archivematica metadata spreadsheets and DSpace Simple Archive Format (SAF) direct to the server filesystem.

As applied to the scanned legacy dissertation project, MAGPIE prepopulates document MARC metadata from a Voyager authority, enables enhancement of the metadata by curators who can read the PDF or extracted text, and facilitates publication into a DSpace Repository. Batch publications can be done with an exported SAF or on an item-by-item basis in the UI with a RESTful push. The Agricultural Research Bulletins have their metadata prepopulated by a provided CSV and further enhanced by providing suggestions via semi-automatic indexing. This project also has the same ability to publish into DSpace. The Spotlight exhibit project is accommodated with the following workflow: MAGPIE prepopulates image metadata via a CSV authority, RESTfully pushes items in batch to Fedora, and allows export CSV for ingest into Spotlight.

In this talk, we will examine how MAGPIE is accommodating rapid growth of our architecture, provide some background on continuing software development and improvements, demonstrate the functionality with multiple repositories and the Spotlight exhibit software, and conclude with the future direction.