Distributed Metadata Creation, or: Let a Thousand Voices Sing!
A colleague of mine in the Art Gallery and I have been working on a project to bring to light about 15,000 images taken by the Soviet news photographer Semyon Fridlyand, who was active in middle part of the 20th century. The task of cataloging these images in the traditional way, that is by professionals, was of course very expensive. Since the collection has tremendous breadth, spanning multiple decades and subjects as diverse as war documentary photography, architecture, farming and agricultural technology, and ethnic dress and customs, it seemed impossible that we could assemble the cataloging expertise to do the collection justice. Faced with the choice of digitizing the photos or cataloging the photos, we chose to digitize them with only minimal metadata–not even a title in most cases, just an identification number.
The images are now available in a discovery system
Collection Discovery Tool
that was developed by a couple of first-rate programmers in the DU Library Systems office. The next step is to develop what we are calling a distributed metadata creation system that will make it possible to open cataloging up to nonprofessional volunteer experts. Inspired by the success of the Steve project, we thought that it was possible to create valuable and valid metadata for our photographs from volunteers. However, we want to take the idea one step beyond simple tagging and develop a system that will create structured metadata–that is metadata that is relevant to particular concepts and “fields” such as “personal name,” or “geographic location.” This way we can use the metadata to populate faceted search and discovery tools.
We are hoping to make contact with people in the former Soviet Union who were the subjects of these photos and engage them in the process of identifying them. In some cases, these photos are the only surviving visual record of this era for these people and places. Suggested metadata will be reviewed by a “panel of experts” who will work within the application to vet and approve metadata and add it to the “official” metadata for the object.
More postings on this exciting project to come.
looks at the intersection of archives, digital libraries, and historical documentation. Greg Colati writes the Quantum Archivist from sunny Centennial, Colorado, where his day job is being the Digital Initiatives Coordinator at the Penrose Library, University of Denver.
This is the missing link! A way to get crowd-sourced metadata into a structured form so it can be brought back into our management systems with less intervention from staff. I can’t wait to hear how this develops!