We developed our own RDF ontology to model our data, based roughly on MODS and MADS, and we store our files and metadata in a custom repository which implements the core of the Fedora 3 REST API. We developed a Hydra head for searching, display, etc.
There is currently an effort underway in the Hydra community called Hydra::Works to build a common data model that can handle complex objects. We plan to implement this model soon using Fedora 4, a Hydra head based on Sufia, and a data model that closely follows DPLA's v4 draft.
If you are coming to C4L in Portland, I will be there there (as will be many other Hydra and Fedora 4 people), and there are also some sessions planned for Thursday and Friday after the conference proper ends.
> On 01/29/15, at 10:25 AM, Sarah Park <[log in to unmask]> wrote:
> Your examples are similar to what I am hoping for. Can you explain a little
> bit more what system you used for backend to store image URLs and Object
> -----Original Message-----
> From: Code for Libraries [mailto:[log in to unmask]] On Behalf Of Esmé
> Sent: Wednesday, January 28, 2015 4:14 PM
> To: [log in to unmask]
> Subject: Re: [CODE4LIB] examples of displays for compound objects and
> At UCSD, we have complex objects which range from a flat list of files (e.g.
> page images):
> all the way up to pretty involved hierarchy modeling a filesystem:
> Many of these have a hierarchy with files attached, but not much metadata
> for the individual parts. But there are also some objects with more
> metadata for each part:
>> On 01/28/15, at 4:43 PM, Laura Buchholz <[log in to unmask]> wrote:
>> We're migrating from CONTENTdm and trying to figure out how to display
>> compound objects (or the things formerly known as compound objects)
>> and metadata for the end user. Can anyone point me to really good
>> examples of displaying items like this, especially where the user can
>> see metadata for parts of the whole? I'm looking more for examples of
>> the layout of all the different components on the page (or pages)
>> rather than specific image viewers. Our new system is homegrown, so we
>> have a lot of flexibility in deciding where things go.
>> We essentially have:
>> -the physical item (multiple files per item of images of text, plain
>> text, pdf) -metadata about the item -possibly metadata about a part of
>> the item (think title/author/subjects for a newspaper article within
>> the whole newspaper issue), of which the titles might be used for
>> navigation through the whole item.
>> I think Hathi Trust has a good example of all these components coming
>> together (except viewing non-title metadata for parts), and I'm
>> curious if there are others. Or do most places just skip
>> creating/displaying any kind of metadata for the parts of the whole?
>> Thanks for any help!
>> Laura Buchholz
>> Digital Assets Specialist
>> Reed College
>> [log in to unmask]