Page 43 - Building Digital Libraries

CHAPTER 3

the library over the years. Reliable workflows and procedures make finding
library resources possible.
The conceptual challenges of incorporating digital collections are simi-
lar to those for incorporating traditional resources. A mechanism to system-
atically identify materials to add to the collection must exist. These materials
must be maintained and delivered in a manner that will be accessible to
users over time and through technology changes. The items must also be
described using metadata or a catalog record. As with physical resources,
some digital resources must be handled individually, while others may be
processed in batches.
Libraries develop a variety of workflows that are optimized to address
the practical challenges presented by different types of resources. Digital
repositories vary significantly in terms of purpose, what they contain, and
the technical and personnel resources available to support them. For this
reason, the processes for selecting, acquiring, processing, and organizing
digital materials vary with the type of resource, as well as with how they will
most likely be used. Although there are many ways to develop workflows,
the most effective approach is often to identify what major tasks need to be
accomplished and then to determine how those tasks can be broken down
into discrete procedures. Once the tasks have been identified, the library
can decide who will do them and how.
Repository design must reflect how people need to use the reposi-
tory—that is, user workflow. When designing a repository, it’s tempting
to construct a data model that supports desired workflows and then add a
user interface to that. However, the front and back ends of systems are inti-
mately related, so using this process can lead to a system that’s optimized for
library rather than user purposes—and this is problematic when specialized
resources are involved.
Instead, the process should begin with an understanding of what the
ideal user experience is, and then a practical plan needs to be devised to
create the data and technical infrastructure necessary to support that use.
Every user experience depends on proper metadata and linkages between
objects, and the success of the repository depends on this process being
realistic and sustainable.
The degree to which automation can simplify workflow depends on the
specific resources and tools available, as well as the technical capabilities
of those using the repository. Even with a very high level of automation,
human intervention will still be necessary to establish certain linkages
between items, whether that involves a controlled vocabulary to identify
related photos, or structural metadata that allows users to traverse a dataset.
The metadata that relate items to each other and which determine how
they can be found and used are what give objects their value—otherwise
no one will be able to use them or know they exist. But creating metadata
requires both time and expertise. It is not realistic to expect content pro-
viders who are not metadata experts to provide this information, nor is it
realistic to expect library staff without domain knowledge to create this
information. Rather, the strategy needs to be to have library staff leverage
28