application profile for images

In the most recent issue of Ariadne there's an interesting article entitled "Towards an Application Profile for Images" by Mick Eadie. JISC has just completed the first phase of some work drafting an application profile for images aimed primarily at the repository community.

I was happy to see recognition of the complexity of digital image objects and the need to track relationships between images, the sources of those images, the content depicted in the images, etc. They looked at FRBR and at the VRA Core, and ended up creating a conceptual model with the digital file at the center that uses the language of FRBR.

I find myself disagreeing with some of their decisions:

In our model, we have renamed the FRBR Work entity as ‘Image’ for reasons of clarity, mainly to avoid confusion between notions of Work as described traditionally in image cataloguing in the cultural sector (i.e. the physical thing) and abstract Work as described in FRBR. As noted above, image as defined in the IAP is a digital image, in line with our notion of end-users searching repositories for digital images of something. Therefore our conceptual model - while still using the language of FRBR and using the areas of SWAP that have applicability across the text and image domains - places the digital image at its centre.

Architecturally, the JISC model makes sense. You have an image, which is a depiction of a site or a work of art, with manifestations as one or more formats of files. That's what you have to physically manage in a repository.

From a discovery point of view, thought, I'm not fully convinced. Having modeled image objects in a cataloging environment and in a repository architecture and discovery environment, I think that FRBR and the VRA Core have it exactly right as to what users are looking for -- images of something. Researchers are looking for images of Chartres Cathedral or Jeff Koons' "Rabbit" or Leonardo da Vinci's "Vetruvian Man." They are looking for images of those Works. When we designed the content models for images in the UVA Digital Collections Repository, the top level is that sense of work. The top level is a work object, which has child expressions for individual views of that work, which have child manifestations that are the actual media files. We found this relatively easy to manage and to build a discovery interface around. It was also the model already used in the cataloging, so it made for an easier translation from that system to the repository.

I need to review the Images Application Profile (IAP) work in more detail. I know this is aimed more at use for IRs and not for image collections, but I think such a profile can only become ubiquitous if it covers both repository scenarios. Their proposed metadata is well-positioned with its use of MIX elements to manage the image files, but I see less than I'd like to see in descriptive elements that support discovery. For example, I don't see a descriptive content date element. I cannot think of a research use case that wouldn't include searching for, say, images of 18th-century French sculpture. I think they should give more thought to incorporating more from VRA Core and CDWA.