Fedora is a general purpose repository system developed jointly by Cornell University Information Science and the University of Virginia Library . The Fedora Project is devoted to the goal of providing open-source repository software and related services to serve as the foundation for many types of information management systems. The Fedora software is available under the terms of the Educational Community License 1.0 (ECL).
The Fedora Project is based on previous research at Cornell University Computer Science that was funded by DARPA and the National Science Foundation.
The Fedora Project is currently supported by generous grants from the Andrew W. Mellon Foundation.
- Powerful digital object model
- The digital objects, or units of information, in Fedora may combine any number and any variety of data streams. These data streams can be local to the repository or may reference content anywhere on the web. For example, one digital object may aggregate a scholarly document in multiple text formats, and another may combine the text, images, and video that are the basis of a rich web page.
- Extensible metadata management
- Because metadata and data are treated uniformly in the digital object model, any number and variety of metadata formats may be stored as data streams, alongside content, in a digital object.
- Expressive inter-object relationships
- Digital objects contain metadata that can express any type of relationships such as membership in collections, structural associations like articles in journals or pictures in albums, or taxonomic relationships. Relationship metadata is indexed and can be searched using semantic web query languages.
- Web service integration
- Fedora fits in with n-tier applications because of two types of web service integration:
- Dynamic content delivery: Web services can be associated with any of the data streams in a digital object. As a result, a digital object can deliver dynamic content: the output of a web service processing data in the digital object. For example, an image processing service can be associated with a digital object that contains an archival tiff image, making it possible to convert the image to other formats (jpeg, gif) or resize and scale the image. Or, a metadata crosswalk service can be associated with a digital object that contains MODS metadata, making it possible to deliver other metadata formats such as Dublin Core.
- Management and Access APIs: A Fedora repository runs as a service within a web server. All of its functionality and all features of its digital object model are accessible through well-defined REST and SOAP interfaces. Thus, a Fedora repository can be easily integrated into a variety of application environments with different user interfaces.
- Version management
- Fedora stores a history of all modifications to digital objects. The full history is accessible through the Fedora access API.
- Configurable security architecture
- Access to all aspects of the Fedora management and access API can be controlled by fine-grained XML-based access-control policies. These policies define sets of rules to permit or deny access by users and groups to repository operations.
- OAI-PMH conformance
- Fedora repositories are fully conformant with the interoperability framework defined by the Open Archives Initiative Protocol for Metadata Harvesting. The Fedora OAI-PMH service exploits Fedoraís extensible metadata management, supporting harvest of any form of metadata delivered by digital objects.
- Preservation worthy
- Fedora repositories incorporate a number of features that facilitate the complex tasks associated with digital preservation. Internally all Fedora digital objects are represented in the file system as files in an open XML format. These XML files include data and metadata for the objects plus relationships to services and other objects. The entire structure of a Fedora repository can be rebuilt from the information in these files. In addition, Fedora repositories are compliant with the Reference Model for an Open Archival Information System (OAIS) due to their ability to ingest and disseminate Submission Information Packages (SIPS) and Dissemination Information Packages (DIPS) in standard container formats such as METS and MPEG-DIDL.