Page 59 - Building Digital Libraries

CHAPTER 3

those identifiers are indexed, since these identifiers will behave like ISBNs or
ISSNs. However, such an approach presumes a separate search for each item
rather than providing a single address that works anytime and anyplace.

Setting Up Workflow

Acquiring, processing, and managing documents requires considerable
human judgment and manual labor, so time, money, and staff effectiveness
can be maximized if tools are optimized to accommodate workflows. Sim-
ply limiting the need for clicks and keystrokes by displaying information
that staff and users interact with in the most prominent positions and by
programming intelligent default system behavior and default values can
significantly simplify and improve processing.
The workflow for acquiring and processing digital resources is often
complicated and labor-intensive. The production and distribution of elec-
tronic resources may be decentralized, so knowing what to acquire can be
problematic. Once materials are identified, they may consist of multiple
interdependent files that require reformatting. Moreover, navigating a digital
resource is inherently more difficult than browsing a physical volume that
has an obvious structure and extent.
Identifying, obtaining, reformatting, cataloging (i.e., creating metadata),
and properly protecting digital resources can be very time-consuming.
However, the value of a repository lies more in how it organizes, presents,
and preserves resources than in how many resources it contains. Therefore,
to reach its full potential, a repository must have both well-conceived work-
flows that support its goals and tools optimized to facilitate those workflows.

Batch Processes

Automating processes can save considerable staff resources and improve the
consistency of the information. However, automation works best for repeti-
tive tasks. Nevertheless, a number of tasks can benefit from at least partial
automation. Automation can easily be used for a variety of tasks such as:
• Identifying materials to include in the repository
• Extracting embedded metadata
• Reformatting or renaming files

• Structuring resources and metadata
• Normalizing dates and other content
• Using application programming interfaces (APIs) to per-
form content analysis on materials
• Creating derivative and archival copies

44