From Files to Objects
When researchers generate data, they typically put their things (notes, files, spreadsheets, directories, etc.) on a storage medium, such as a hard drive. So how do these files end up as digital objects in a collection?
Intellectual and technological processes transform all that 'stuff', mostly files and information, into digital objects. A large part of the intellectual work is in organizing and describing that stuff, or creating metadata.
The Metadata Process
For each research data curation project that has been agreed to be stored in the UC San Diego Digital Collections, there are metadata consultations and transactions at various points in the project lifecycle. When metadata has reached a mature state, the Digital Object Metadata Management team then crosswalks or maps the metadata to align with the data model of the Digital Collections' Digital Asset Management System, or DAMS. Meanwhile, the files and data for the project are ingested into the DAMS, and after the metadata is mapped, associated with their objects. The collection of objects (which you'll remember is comprised of files and metadata) is then built. After quality control by metadata professionals and collection stakeholders, and given that there are no embargoes or special access limitations on the collection, it is ready to be viewed at the Digital Collections website: UC San Diego Digital Collections.
Metadata Standards Used in the DAMS
The current iteration of the UC San Diego Digital Collections' metadata follows the RDF data model, and is serialized as RDF/XML, which as the name implies, is RDF expressed in XML. This means that utlimately metadata is stored as a series of XML 'triples' in the DAMS Triplestore. A triple is simply a statement, using the framework of a natural language assertion: 'Subject - Predicate - Object'. These statements aid computers in defining and establishing relationships between entities and objects, which is one of the goals of Linked Data. Future work for DAMS metadata will be to facilitate better connextions to the Web of Data.
The DAMS Data Model
The UC San Diego Digital Asset Management System, or DAMS, is in its fourth iteration, referred to internally as DAMS4. The goals for the DAMS4 Data Model are to:
- Be flexible to accommodate disparate metadata from a variety of sources
- Promote consistency within the data store
- Support relationships within and between objects
- Engage the community by sharing vocabularies and technology
- Support improved management of objects and metadata
Digital objects themselves are not merely files, but are composed of other 'components', files, and metadata in a hierarchical manner. For example: