Learning from the SCI-Arc Media Archive

The expertise and resources required to create accurate and useful textual guides for a large collection of moving image documents exceed the means of most institutions. However, during 2011-2012, when the Southern California Institute of Architecture created an online Media Archive of lecture videos (http://sma.sciarc.edu), it devised a process for harnessing the subject expertise of non-professional cataloguers to generate searchable descriptive texts for 1000+ hours of video in three months. The paper outlines the creation of an efficient online work interface, plus a simple step-by-step procedure that would minimize the need for video or cataloguing expertise. The creation of three kinds of searchable texts are discussed: controlled vocabulary names and verified dates, narrative descriptions, and brief subject tags known as “Themes.” The narrative descriptions were intended to capture the idiosyncratic vocabulary of each speaker in its specificity, while the keywords were more generic, collocating diverse expressions of a similar idea. The paper also discusses the drawbacks of this approach, including the protracted training period at the beginning, and the intensive copy editing and fact checking required at the end. Hopefully, a concise outline of our strategy and process will encourage others to re-examine their own video resources—and act to save them before it is too late