Where does the DS data come from? How is it created?
In order to build our national union catalog, Digital Scriptorium takes whatever structured data a member institution has relating to its manuscript holdings, usually in the form of institutional catalog records. We aggregate and harmonize this data based on our data model, which conceptualizes the metadata (catalog) record as independent from but linked to the physical manuscript object in the real world. Administrative metadata regarding holding information is also independent from but connected to the physical manuscript, creating a backbone of three core entities: DS Record, Manuscript, and Holding Information. This allows us to establish a unique and persistent identifier for the Manuscript object, while offering the flexibility to modify catalog records and metadata descriptions about the object (DS Record) as well as to maintain and update holding information, both current and historical.
Data is contributed by member institutions in whatever encoding formats they use locally. Contributed data is first processed by mapping member metadata to our data model. We then extract data values based on metadata element type and enrich those data values, where possible, with existing Linked Open Data (LOD) authorities and vocabularies. This semantic enrichment process allows data strings to be meaningfully linked to their LOD equivalents, giving greater context to member data, which will help render it into LOD. (To learn more about our semantic enrichment process through OpenRefine, please see our GitHub documentation.)
How is the value of DS data enhanced by Linked Open Data? Why does it matter that DS data is enriched by LOD?
LOD is all about providing open access and rich context. When data is made publicly available in a format that allows it to be meaningfully linked with other data, the value of that data is increased. Publishing data as LOD through semantic enrichment is an increasingly successful strategy for creating and maintaining smart data in the digital humanities.
Keeping smart data principles in mind, the value of member data is increased when enriched as LOD uploaded into our Wikibase, a knowledge graph in which member data is aggregated, published, and stored as LOD. Our data model supports representation of original member institutional data in addition to enhancing it with internal and external LOD authorities and vocabularies. This renders DS data as interoperable with other LOD data using or connected to the same vocabularies. DS data can also be queried using the Wikibase Query Service, a SPARQL endpoint for our DS Catalog.
How and why is the new DS Catalog different from the old one?
The primary difference between the new DS Catalog and the old one is that DS no longer manages the creation and maintenance of data, including images, but instead relies on member institutions to manage and contribute the best data possible to the DS Catalog. You can learn more about the DS 2.0 Project here.
Why are some records more detailed than others? Why do some records have images and others don’t?
The level of data contributed to DS by member institutions depends on the resources for creating and managing data at each institution. DS requires only that a member institution recognize that it has a manuscript to be included in the DS Catalog (i.e. provide holding information). Some institutions have more resources and can produce expert-level descriptions and full digitization. Other institutions may not have the ability to commit the same level of resources but can at least provide some data in order for their collections to be counted among DS manuscripts. We hope that for these institutions, the DS community of researchers, scholars, and students will share knowledge to help institutions improve their manuscript descriptions.
What happened to the old data?
DS has preserved copies of the original DS Catalog metadata and images in its Zenodo Community and on the Internet Archive. More information documenting the process and results of the DS 2.0 Project can be found on our Research and Documentation page.
If I have more information about a manuscript or want to suggest a correction to the data, how do I share that information with DS?
DS Catalog users are encouraged to contact member institutions directly about suggestions to improve the DS data. DS does not correct the member institutions’ provided data, but does perform regular updates in cooperation with member institutions. Each record has a link to the institutional record where you will also be able to find the best contact info. You may also visit member institutions’ profiles in the Member Directory for contact information.
Do you have other questions, concerns, or suggestions? Let us know your thoughts by completing this survey.