Information Technology Reference
In-Depth Information
Understanding our users
The change in user behaviour as users move towards a fully digital
library brings about a real possibility that librarians and libraries will
become removed from their users (Nicholas, 2008). However, this
change in behaviour also creates an opportunity to collect user
information and to monitor use; Walker (2009) likens this to being
seated on a goldmine.
Text and data mining
Unlike the physical library, where little objective information could be
gleaned from irregular usage checks, it is no longer satisfactory to think
that we know what is best for the user. We can utilize usage statistics
to inform us of the potential value of a given resource, but we can also
look much deeper into how users behave: which sites do they refer from,
how long do they stay, what do they do when they are there, where do
they go? This allows librarians to make informed decisions (CIBER,
2007). It does not mean to say that we cannot select or recommend
what the user may need, but we must do it in an informed way, and in
their language.
At the University of Huddersfield, Web 2.0-style recommender
services and usage logs are being used to create additional features on
the library catalogue to enhance resource discovery (Pattern, 2009a).
Although some of this functionality refers to both electronic and print
resources, it should be noted that for many subjects print is (and will
continue to be) a major source of much of the scholarly content and is
therefore crucial to resource discovery, even in a digital world:
Zero hits
: accounted for 23% of searches over a six-month period; this
now automatically generates a 'did you mean . . .', '. . . or maybe'
spell checker and 'serendipity searches'. The serendipity search
generates suggestions by running the search against a number of
external reference websites (including www.answers.com). The
results returned from these sites are then compared against the
catalogue to generate a series of potentially relevant keyword
searches. Users are then invited to use their own judgement to
decide whether or not the suggested searches (and also the results
they generate) are relevant to the original search.
