The many faces of digital archives

This past week I continued to work on the Los Angeles Examiner photo archive. I noticed that in the early 1950's "Commies" were in the news, as were Korean War soldiers and families.

Happily this work was interspersed with meetings! They apologize about all the meetings, but it's nothing new for me, except that I can attend as in interested by-stander, without responsibility. The CIS (Collection Information System) group is project based, charged with moving the existing (legacy) digital archive to the new platform based on Documentum. Their weekly meetings, conducted by the project manager Tim Stanton, serve to make sure each subgroup is on task and to handle any problems along the way.

The subgroups work on thesaurus management, metadata and migration of it to the new system, the web user interface and testing thereof for usability, the Contributor Module for ingesting new documents into the system by staff and/or faculty themselves, and production (modifying the system architecture and documenting it). These meetings give me a chance to hear input from these various components of the project.

I notice that process is almost as important as content. The focus on project management is intense, with deadlines strongly adhered to and reporting required frequently. This culture is fairly new I am told, replacing a former casual culture that did not always produce timely results. While the goal of project management is noble and probably necessary given the complexity and size of the project, I feel a certain tension between time demanded for process versus time needed to "do the work."

Another meeting was DIM, Digital Imaging and Management. These are the people I am working with right now. They meet to discuss status of imaging projects, metadata, migration of metadata to the new system, needs of new projects coming into the system, etc.

I had occasion to see one new project firsthand. Friday morning Wayne and I met with people at the Seaver Center of Western History Research at the Natural History Museum south of Exposition Boulevard. I had never been in the building. It's a great museum; I hope to get back there as a tourist. At the Seaver Center they are digitizing photos from various collections depicting western history. Then they will be added to the USC Digital Archive. USC already has received some of the images, but the topic Friday was the descriptive data and how it would be migrated from the Center's Excel spreadsheets to the Archive. Wayne had prepared a sample of records showing conversion from an Excel record to a record in Dublin Core (with all the descriptive tags) as it would appear in the USC Digital Archive public interface. He then reviewed the intermediate steps of mapping each Excel column to a Dublin Core element such as title, subject, date, etc. And finally he showed how the record would look in XML for import into the CIS Oracle system.

The Seaver Center staff were having to make decisions about number of subject headings, use of synonyms (include as subject headings? as part of descriptions? or just let future thesaurus modules create the synonyms?), whether size of glass plate negatives or photos was something users needed to search on or sort by, forms of names, etc. They had to decide whether to allow several objects (photo, negative, glass photonegative) to be on the same record, or whether each object needed its own record. There was also discussion about ID numbers and whether a user could recognize the appropriate ID or "call" number needed to enable staff to find the originals if necessary. They also had to discuss copyright notice and whether a statement with each image or one for the collection would be sufficient. There are many details involved in making sure the final records will look good, be searchable, and provide staff as well as users with sufficient information about each record.

So it was a full week of learning. I am still sifting through it all, trying to make sure I understand how everything fits together. It is confusing because they are dealing with two systems, the legacy BRS-based home-grown system and the new Documentum-based system. The initial Seaver collections will come in under the legacy system to move the project along, rather than wait for the new system to be ready. Migration to the new system will take place later.