Managing Different Types of Digital Object:

VIF research shows that the majority of repositories either already accept a variety of types of digital object, or are considering introducing other types in the near future. 

These objects are in some ways harder to deal with than text documents because there are more variations in the sort of versions associated with them, such as formatting and compression, and they can have complex version relationships with other objects.

The majority of the guidance in this framework relates to all digital object types and where there are differences these are highlighted.  However, when considering the guidance, and in particular which versioning solutions to implement, do consider what would work best:

Previous research into using repository platforms for 'multimedia' objects has shown this to be a much more immature area of practise than text based repositories. The MIDESS project (external link) gives an interesting insight into the complexity of setting up a resource containing images and multimedia items.

Images, Video and Audio:

The objects are usually made available as compressed versions of an original file. They are also often converted into different file formats depending on their intended use. Detail versions of a picture, where a part of an original has been extracted/cropped and perhaps enlarged are also common, for example. These objects are clearly versions of the original, and this relationship needs to be expressed in either in the object, repository record or the metadata held about the object. The MIDESS project (external link) investigated preservation issues for multimedia items in a recent report (pdf).

Specific to images and of interest to the versions issue is the Digital Images Archiving Study (external link) from the Arts and Humanities Data Service (AHDS) (external link). They recommend preserving images in multiple file formats, therefore creating multiple versions of each image by default.


VIF research showed that deposit of data is currently uncommon in non subject specific institutional repositories. However, demand is growing for tools to facilitate data storage, and it is an area of huge potential growth in the future. Please examine the following work in this field:

The metadata requirements for different sets of data can vary greatly and are different to metadata usually kept for text documents. New initiatives are working on metadata profiles, such as:

Learning Objects:

Learning objects are rarely held in institutional repositories used for open access purposes, and are more commonly in specific resources such as Jorum (external link). Versioning of these objects is quite different to that of the other object types listed above, because as learning objects are refined or altered for a specific course or purpose, older versions become irrelevant. Jorum currently only make the most recent version of a learning object available (this position is being reviewed in early 2008). However, older versions are archived within the system should they be needed for reverence in the future, and Jorum staff allow the object creator to decide whether any older versions should be made available.

This is in contrast to the common practise in repositories of resisting representations from academics about which versions should or shouldn't be made available once they have been deposited in a repository.

Related work on the future of electronic learning can be found in Charlesworth, A and Ferguson, N and Schmoller, S and Smith, N and Tice, R 'Sharing eLearning Content: a synthesis and commentary', is available here:

Metadata for Different Object Types:

It should also be noted that several metadata application profiles are recently in development, each expanding simple Dublin Core metadata to allow for richer description of different digital object types. See the the Application Profiles section for more detail.


The next area of the Framework is information for Content Creators.

Last updated 14/05/08 | Copyright © 2008 LSE