descriptive metadata. This is the metadata that is transcribed from the
piece being described. In this case, descriptive metadata included the
song title, first line of the refrain, and publication information. I also
showed the students how to search the Library of Congress’s free
online authority file (Library of Congress, 2009) for authorized forms
of names for the lyricists, composers, arrangers, performers, and/or
illustrators credited on the pieces. Finally, I created standard notes for
the student assistants to apply, such as “piano, vocal” for an
instrumentation note, “One color (purple)” describing the cover art,
and “Includes advertisements” as a miscellaneous note. I reviewed
their work and completed the metadata with subject analysis, detailed
cover art description, and additional notes.
Learning how to use CONTENTdm required a tremendous
amount of time and energy during this first project. The effort was
amply rewarded, however, since the functionality provided by the
software suited our application perfectly. The software supports batch
population of a collection via tab-delimited files. This facilitated
collaborative metadata creation, since the graduate student workers
could create Excel spreadsheets with preliminary metadata for groups
of titles and then pass them on to me to complete. I subsequently
converted the spreadsheets into tab-delimited files and uploaded the
metadata into CONTENTdm along with the corresponding images.
The compound object structure, in which several images comprise one
digital entity, elegantly matches the character of multi-page sheet
music. The software also provides the means for creating index boxes,
which enhance access to the content beyond full text searches of the
metadata. We decided to use this functionality to create index boxes
for composers, lyricists, and subjects for this project.
As stated previously, Lovejoy Library’s digital collections are
created under the consortial umbrella, using CARLI’s CONTENTdm
server. CARLI’s collection of member libraries’ digital collections is
OAI-harvestable, and CARLI provides a means for member libraries
to obtain usage statistics. But with these advantages come some
constraints. CARLI requires all of their hosted collections to contain
certain metadata fields, including (among others) Rights and
Language fields mapped to the corresponding Dublin Core elements