Enhanced Content Models for Fedora

Show simple item record

dc.contributor.author Blekinge-Rasmussen, Asger
dc.coverage.temporal Date: 2009-05-21 08:30 AM – 10:00 AM
dc.date.accessioned 2009-06-12T20:24:24Z
dc.date.available 2009-06-12T20:24:24Z
dc.date.issued 2009-05-21
dc.identifier.uri http://hdl.handle.net/1853/28531
dc.description 4th International Conference on Open Repositories en
dc.description This presentation was part of the session : Fedora User Group Presentations
dc.description Date: 2009-05-21 08:30 AM – 10:00 AM
dc.description.abstract This presentation introduces Enhanced Content Models for Fedora. Enhanced Content Models have a number of new features compared to the Fedora 3.0 content models. First of these is the more elaborate specification of the data objects. Second is the repository view system, which allows the repository to dynamically remap the contained data to virtual data objects. And third is the object creation templates, which allows the content models to behave as object classes from which new data object instances can be made. All our work is under the Apache 2.0 License, and will be available as add-ons to Fedora. Fedora is an extensible repository system, containing data objects and content models, which hold descriptions of the data objects that subscribe to them. In Fedora 3.0 Content Models express the classes of objects, and tie data objects to disseminators, but do little else. Content Models are formal descriptions of data objects, which should be distinguished from datamodels, which are descriptions of collections of data. Having a datamodel is a requirement for many digital repositories and the easiest solution is creating an interface that only allows data to be entered in a special format. If all data is input through this interface, it will adhere to the datamodel. Unfortunately, this has the side effect of coupling the datamodel to the program code of the interface. We believe that the datamodel should not be part of the interface, it should be part of the repository. We achieved this by enhancing the Content Models. The Enhanced Content Models can specify the cardinality and target classes of relations, and schemas for datastreams. We have implemented a validator, which checks data objects against their Content Models. A set of Enhanced Content Models makes up a datamodel. In Fedora, you might have atomic objects making up a "record", but for indexing purposes, this record must be flattened to one compound. Enhanced Content Models can specify how to do this flattening, in the repository view system, and we have implemented a webservice to create such compounds. In many OO programming languages new objects are created as instances of a class. Enhanced Content Models implements this pattern. You can declare certain data objects to be templates for an enhanced content model. We have developed a webservice that can create new objects in Fedora, given a content model and a template to use as basis. en
dc.description.sponsorship Denmark's Electronic Research Library en
dc.language.iso en_US en
dc.publisher Georgia Institute of Technology en
dc.relation.ispartofseries OR09. Fedora User Group Presentations en
dc.subject Content models en
dc.subject Datamodel en
dc.subject Enhanced content models en
dc.subject Fedora en
dc.subject Ontologies en
dc.subject Schema en
dc.subject Validator en
dc.title Enhanced Content Models for Fedora en
dc.type Proceedings en
dc.contributor.corporatename Statsbiblioteket i Århus


Files in this item

Files Size Format View Description
FedoraOR09.ppt 211.5Kb Microsoft PowerPoint View/ Open Power Point Presentation
FedoraOR09.pdf 166.7Kb PDF View/ Open PDF Presentation

This item appears in the following Collection(s)

Show simple item record