Page 53 - Building Digital Libraries

CHAPTER 3

entries in the repository. An obvious solution is to create a record or index
page listing the issues in reverse chronological order even if the original
publisher provides no similar navigational tool.
Although electronic serials can be made navigable with the appropriate
metadata, such a solution requires the system to properly interpret and sort
a wide variety of enumeration schemes. These various enumeration patterns
would invariably need to be entered as plain text because there are too many
different patterns for a system to list them all. Also, the serial title would
need to be entered exactly the same way in each record. Otherwise, even
minor discrepancies would cause different issues to appear under different
titles—which would make it effectively impossible to browse issues.
Likewise, digital resources are even more likely than physical resources
to consist of many individual components. With paper materials, the
indexes, appendixes, and other supplementary materials are usually bound
with the main work. However, digital resources might store different sec-
tions of text and supplementary material in separate files, particularly when
the supplementary material is in a different format from that of the main
document. It’s possible to express the relationships between the various
components with metadata, but it’s very difficult to do this in a way that will
allow the work to be used as an integrated whole. As is the case with serials,
the slightest typographical error can separate the components from each
other. Again, to make the resource usable, a separate record aggregating the
resources or even an index page listing the components in an appropriate
order might be far more helpful than complex metadata.
Despite the challenges it presents, good-quality metadata is usually the
core of any good organizational scheme. Metadata can store information
about the format, purpose, associated people and organizations, time, place,
relationships with other resources or collections, and any other appropri-
ate information. It is important to be aware that the system has to be able
to interpret the metadata for it to be useful—there is little if any value in
storing information that systems do not use. With few exceptions, storing
metadata that current systems cannot use is unproductive.
For example, over the years, catalogers have spent enormous amounts of
time meticulously encoding cryptic fixed fields that store information about
various types of illustrations and tape formats, as well as whether the item
contains conference proceedings, is a Festschrift, or contains bibliographi-
cal references. No major integrated library system (ILS) uses any of these
fields or a number of others that are equally obscure. Most of these fields
are ignored outright, and there is no patron demand to see or search them.
Consequently, it is unlikely that this information will ever be usable. It is
not difficult to imagine scenarios in which some obscure bit of information
would be useful. However, the fact that someone might conceivably find
a use for the information does not justify storing it. Storing information
“just in case” someone may need it at a later date takes resources that would
probably be better spent providing information and services that people
already need.

38