And just to clarify Sandy's examples, here are the content models that we
have developed so far (more to come as we look at video, audio, and data
sets) at UVA:
uvaBitonal (1-bit static preview, screen size, and Mr.Sid)
uvaLowRes (24-bit static preview and screen size; no Mr.Sid)
uvaHighRes (24-bit static preview, screen size, and Mr.Sid)
uvaGDMS (xml metadata for images)
Electronic Texts (TEI):
uvaBook (xml metadata header & transcription and page images)
uvaGenText (xml metadata header & transcription with no page images)
uvaPageBook (page images with xml metadata header but no
Electronic Finding Aids (EAD)
uvaEAD (xml finding aid, e.g., metadata and full text description
of a collection)
An image object can be a child of image metadata, electronic text, or EAD
At 03:03 PM 3/10/2005, Sandy Payette wrote:
>In the most general sense a content model is intended to describe a
>particular design pattern within the context of the Fedora object model.
>What I mean by this is the pattern of datastreams and disseminators as
>related to a certain type of content entity. For example, a Fedora digital
>object that represents an electronic thesis may have one datastream that is
>metadata about the thesis (of type text/xml), and one datastream that is the
>thesis content (application/pdf). It may also have a disseminator that
>converts the pdf to another format. This whole pattern would be identified
>as a content model, and it would have its own unique identifier (ideally a
>Currently, we are working on how to formally specify content models. We want
>to come up with a way to encode a content model specification (ideally
>XML-based). A formally specified content model can enable extended
>validation of digital objects (i.e., does this object conform to content
>model A?). This is different from validating a Fedora digital object as a
>generic entity (i.e., is this a valid Fedora digital object?). Another use
>of a formally specified content model would be to make configurable client
>applications for creating digital objects (e.g., the user can select a
>content model template and the right inputs would be rendered in a client
>interface to assist the user in building the right kind of object.)
>This month a group of Fedora collaborators will be conventing to work on the
>content model formalization problem.
>Up until this point in time, people have been specifying content models more
>informally in their Fedora implementations (via spec documents, best
>practices, and custom validation). They have been putting a simple
>identifier in each digial object to indicate which model the object belongs
>to. But, it's really just like assigning the object to some group. For
>example, at University of Virginia, they have a standard image model
>(UVA-STD-IMAGE), an EAD model (UVA-EAD), and some other models. At UVA,
>their front-end workflow utilities create Fedora ingest files that conform to
>Hope this helps for now. Tutorial #3 will be written after we finalize the
>formalization of content models.
> >>-----Original Message-----
> >>From: fedora-users-bounces@...
> >>[mailto:fedora-users-bounces@...] On Behalf Of
> >>Tomislav Karafilov
> >>Sent: Wednesday, March 09, 2005 10:24 AM
> >>To: fedora-users@...
> >>Subject: [Fedora-users] Tutorial 3: "Fedora Content Models"
> >>completitiondate ?
> >>can someone from the Fedora Development Team tell me, when
> >>the Tutorial 3 estimatedly would be finished and/or what is
> >>meant with "Content Model" in Fedora ? Is the implementation
> >>for "Content Models" already done in Fedora 2.0 ?
> >>Tomislav Karafilov
> >>Fedora-users mailing list
>Fedora-users mailing list
Director, Digital Access Services
University of Virginia Library