Page 223 - Building Digital Libraries

CHAPTER 9

often this ranking is user-initiated, and done after the items
have been queried. Given the inherent speed limitations built
into a broadcast search system, federated systems traditionally
print results as they are returned from target databases, mean-
ing that unlike traditional search engines which provide the
most relevant items at the beginning of a results set, federated
search systems require some user interaction to replicate that
result. What’s more, given the varied nature of the data returned
by target resources, ranking items within a larger result set can
be a challenge in and of itself, and this continues to be fertile
ground for further research and discovery.

Development of language agents
Cultural heritage organizations’ primary research and user
communities are more diverse and international than ever
before. This diversity has significant benefits, since new users
are approaching the evaluation of primary digital content from
a wider cultural lens. At the same time, it has exposed some
significant challenges, particularly in the United States, where
materials and metadata are overwhelmingly available only in
English. This lack of language diversity has a major impact on
the findability of content for users, creating significant barriers
to access for many within the international community.
To address these challenges, research into the development of language
agents seeks to build a bridge between users and systems, and to overcome
traditional language barriers. This work is possible due to the continued
improvement of the real-time translation services that are being used to
power personal digital assistants. Both Google and Microsoft are investing
significant resources in the development of artificial intelligence services
to power their personal digital assistants. Researchers interested in lan-
guage development have benefited from this work, since both Google and
Microsoft provide open web services to interact with these systems. This has
enabled researchers to delve into the creation of intelligent agents that could
be used to improve the library discovery experience by removing language
as a barrier to discovery. The work is still in its initial stages, but this seems
like fertile ground for continued development.

Searching Protocols

As the library community has become more dependent on new and emerg-
ing technologies to deliver and maintain services, so too have libraries
found themselves becoming more dependent on specific technical metadata
standards and protocols. Chapters 5 and 6 discussed the library commu-
nity’s current reliance on XML and XML-based metadata schemas within

208