Page 220 - Building Digital Libraries

Thinking about Discovery

to control how outside resources interact with its local tools. This can give
institutions the ability to create or support standard web services’ APIs to
enable better integration with their federated search software. Likewise,
local digital projects are more likely to accommodate local data harvest-
ing, enabling tools to harvest collections into a single repository for faster
searching.
Digital repository programs can also take advantage of a federated
search program to expand their acquisition and collection development
resources. Many digital repository developers will often make the mistake
of considering only their organization’s digital content and projects as
“collectable” digital resources. However, a good federated search program
allows an organization to fully utilize digital collections from not just its own
digital resources but digital resources from other organizations, removing
the barriers of organizational ownership and distance for its patrons. This
allows users at one organization to query resources from not only their
institution, but from selected digital collections held by other institutions.
In this way, an organization could potentially leverage the Library of Con-
gress’s American Memory Project and OAIster alongside local and vended
content. Finally, federated search tools offer an organization the ability to
provide value-added services for its users. Federated search tools can allow
organizations to capture search history and document click-counts in order
to augment their ranking algorithms and provide context to search results.

Why Think about Discovery

Given the improvements made by search engines like Google and Bing, one
is left to wonder if organizations still need to provide their own discovery
systems. Within the library community, discovery tools have traditionally
been used to link vended content with other resources—which to some
degree is starting to happen at the search engine level through Google
Scholar and other aggregated research-focused portals. Given the devel-
opment of these portals, should local discovery even be a consideration
for digital library administrators? The answer would be a resounding yes.
While traditional search engines continue to do a better job of indexing and
surfacing academic content, the reality is that large swaths of digital content
remain outside of their view. Whether this information is vended content
purchased by the library or is in locally managed digital collections, organi-
zations must make an active effort to ensure that content is discoverable by
these tools—and even then, there is no guarantee that one’s materials will
show up when queried. Certainly, organizations should be actively working
to ensure that their content is accessible and indexable within the traditional
search engine infrastructure—but these systems utilized closed relevancy
algorithms that make it difficult to know how one’s content will be searched.
The best way to ensure that local content remains findable is to create a

205