Searchable by Others

Trudy Levy

Have you reached the next stage of managing a digital image collection?  This is the stage where your superiors acknowledge your success by asking you to now expand the system for the whole institution.  This is your golden opportunity, but you do not know their search habits or for what purposes they want to use images. You do not know how to catalog images so they are searchable by others.

Two of my clients have brought this problem to me.  One, the California Academy of Science, had created a successful slide collection using Cumulus. The collection contained scientists’ field images.  Now they not only wanted to add other collection, but also make it accessible to several more departments.  The other, Los Angeles Metropolitan Transportation Authority, wanted to create a common image collection and distribution system for three divisions within the Creative Services group.  In the beginning, I thought it would be a cataloging problem, but eventually I realized that searching and cataloging did not necessarily correlate on a one to one basis.

You must catalog an object in order to find it again, but you also catalog to relate relevant information to the image.  According to whatever standard a cataloger chooses to use, the related data is pertinent to fully understanding and using that image or file.  It may be technical data describing the actual file such as image size, or administrative data such as file location or creation date.  One would hope there would also be right data associated with the file and of course the broad category of descriptive data.

However, the likelihood of anyone searching for an image based on any of that data is slim.  It is not that we have all become googlized either. When people are looking for an image, most of the time they want it for its communication power.  Whether it is to reinforce a point in a lecture or to motivate people, people are looking for those thousand words that the image is worth.

Once I realized that, solving the two organization’s problem became relatively easy.  Each collection would continue to be catalog according to the standards developed by their own authorities.  Thus, at the California Academy of Science the field scientists would organize their images according biological geographical taxonomy.  The archivists would organize their images of the historical photographs according to good archival practice.  The exhibit people are just starting to create their system and the research scientist are participating in the development of a global biodiversity data structure.  So at the MTA, each division would assign the data to their images that they felt was necessary for good and accurate usage. As most DAMs can search across several collections, there is no need for the different collections to have same system but do they need the same search fields for general searching.

It was at the MTA that I discovered the secret to finding the right data required for general searching. At the Academy, we had worked as a team to develop a management system that could manage multiple collections, while maintaining the individuality of each collection.  At the MTA, I changed this approach to one that created a central collection being fed by the individual collections.   This is a subtle change, but it made life much simpler.  I created this approach because ironically at the MTA they did not want to share everything. At the Academy, on the other hand, even though they had to maintain different collection structures due to the complexity of their collections, their sense of collegiality led them to want to share it all.  Each curator wanted people to see all their data, but we found that the quantity of information just baffled the general or even specific searcher.  In fact, even the curators admitted to using keyword or category searches.

At the MTA, we did not try to work as a team.  This goes against my personal work ethic, which is to bring people together, but I have to admit this worked.  I interviewed each group separately in terms of what information they currently associated with their image and how they searched for them.  More importantly, I ascertained WHY they would look for an image.  I identified the common fields and terms, including a few must haves, and presented this data to the total group.  We started this project with a search for the commonality and were able to find it through individual effort.  The key was recognizing the importance and value of each collection’s data structure for assuring proper use and at the same time accepting that search might be answering a different question. 

August 22, 2004