Basic required elements
Each item submitted to DPLA must have the following metadata:
- Unique ID
- Title
- Link to full image
- Link to thumbnail image
- Rights Statement
- Providing institution
More technical information is below; if you’re not sure about next steps, contact us.
Remediation and Modelling
Metadata must be modeled consistently across the Digital Virginias Hub and aggregated by the hub administrator (University of Virginia) before it can be harvested by DPLA. To help with this task, contributors use a custom version of the Qualified Dublin Core term set to model description for all aggregated resources. After metadata is harvested, it is mapped to the DPLA’s Metadata Application Profile (DPLAMAP) for display in their portal.
Remediation is an important step in the data submission process where metadata must be carefully assessed for quality and consistency before aggregation. This includes things like eliminating blank elements, defining and applying elements consistently across collections, using controlled vocabularies where possible, standardizing date values, and omitting spelling or punctuation errors.
Additional tools are available for assessing and transforming data per the requirements listed here. Contact us for assistance or if you have questions.
Aggregation and Harvest
Digital Virginias uses the git software and the hosted Github service to manage metadata aggregation and harvest. Git is a distributed version control system that is ideal for tracking changes to large sets of text-based files. GitHub is a commercial service that centrally hosts git repositories and offers a wealth of features to enable collaborative development.
Each hub partner is assigned a repository at the Digital Virginias GitHub profile, which they have access to through their Github credentials. Partners can then push new metadata records and revisions whenever they prefer. Once content is pushed to Github, specialized software automatically runs a validation service that checks for baseline metadata quality requirements as defined in our local XML Schema. If any files are identified as invalid, information is made available via visual status and linked to in a report in Github. If all records are valid, a branch repository will reference that point in the commit history, which can be accessed for harvest by DPLA.
Additional Resources
- To learn more about the Digital Virginias metadata element set, consult this crosswalk which lists elements currently available, requirements, and mappings to the DPLAMAP.
- DPLA provides some very helpful documentation including a specification for the DPLAMAP as well as an introductory document that explains the model.
- Metadata experts from Digital Virginias partner institutions have collaborated to provide an introduction to workflows and a baseline set of guidelines for modeling and preparing metadata to be harvested. These guidelines can be consulted as a compliment to your preferred content standard and the requirements specified in the DPLAMAP.