Fedora and Mulgara

The Ensemble Project uses Fedora Digital Repository to enable researchers, teachers and students involved in different settings to create and manage collections of datasets from various disciplines. An important part of the research carried out in the technology strand has been the analysis of several digital repositories including DSpace and Fedora, among others. Fedora was selected as the basis of the applications built by the project in collaboration with teachers and students. Since different settings present different requirements and use data differently, we need a solution flexible enough to cover all of these requirements.

Some of the developments within the different settings include the following uses of Fedora repository in combination with Mulgara:

  • Access to teaching/learning resources provided by teachers and students. Depending on the nature of the applications being used in teaching environments, Fedora might be simply used as a normal content management system, holding teaching materials collections, which can then be accessed by teachers and students. In this respect, Fedora provides a very flexible framework that enables structuring of collections of resources with support for a wide range of resource formats: audio, video, texts, images and even XML/RDF data.
  • Virtual Collections. Some settings use the repository to store references to resources accessible from external providers such as publisher’s sites or other sources of information. In this case, the repository is used to manage those collections of resource references accompanied by descriptive metadata annotations.
  • Data aggregation and semantic applications. This scenario is the most complex but also the richest and uses the semantic capabilities of the repository combined with the triplestore facilitating aggregation of data coming from multiple sources. An example of this type of use is Outdoor and Environmental Education where students and teachers work with geographical data obtained live from websites or constructed from fieldwork. The data is included in the triplestore and can then be queried and presented using implemented web applications.

Fedora

The Fedora Repository (Flexible Extensible Digital Object Repository Architecture) is capable of serving as a digital content repository for a wide variety of uses. Among these are digital resources management, institutional repositories, digital archives, content management systems, scholarly publishing enterprises, and digital libraries.

Fedora offers many useful features such as the new digital object model that allows to store, within an object, metadata annotations using different schemas, data in differing formats and even semantic-ready data in the form of internal RDF/XML data. Fedora also provides a powerful set of tools for data ingest, description and management, and it is close-coupled with the Mulgara triplestore. The latter provides access to data and metadata from the digital repository, which can then be combined with data from other sources. Additionally, rules and inferencing can be applied and query interfaces used to populate web interfaces or lightweight visualization frameworks like ‘Exhibit’ from the MIT Simile toolkit.

Mulgara Triplestore

Mulgara is a semantic triplestore database that holds data in the form of subject-predicate-object statements extracted from data in various formats such as RDF-XML, n3, RDFa among others. Examples of the type of information that could be stored include Dublin Core annotations for bibliographical reference:

http://someurl/mypublication dc:creator A. Parker

Since we are working with data coming from multiple sources, e.g. a repository, live sites like Wikipedia, publishers’ sites etc., we have decided to use a separate Mulgara server. Fedora is then configured to access our Mulgara instance instead of using the embedded instance that comes pre-installed with the repository.

On the other hand, Mulgara can be used as a remote web server and it provides several mechanisms/interfaces to access it

  • HTTP SPARQL and iTQL Web Services
  • Web user interface to make iTQL/SPARQL queries
  • SOAP endpoints that can be used within other applications using this mechanism to access the Mulgara database
  • Java API’s including JRDF (uses the standard Java RDF libraries to query and use the triplestore) and Jena compatible methods (allowing connection to and use of Mulgara using Jena libraries as well as its reasoning capabilities)

More information about Fedora

More information about Mulgara

Read more in our publications