Finding Your Own Assets: It Takes More than Just Having a DAM

Page 1 of 3

May 02, 2006

May 2006 Issue

      Bookmark and Share

As digital asset collections grow, it becomes increasingly important to build a digital asset management system (DAM) with strong classification, taxonomy, and search components to help you locate an asset whenever you need it.

Yet digital assets present a unique search problem because the search is really only as good as the metadata and keywords you have associated with an item. With a text document in a content management system, even if the title is not terribly descriptive, you can always rely on a full-text search to bail you out. With a picture or video, you don't have that luxury. Therefore, when you build a DAM, it is imperative that you carefully craft a detailed taxonomy and classification system with strong and meaningful metadata and keywords, or your assets will be buried in the system, lost from the users who need access to them.

Part of the problem with trying to store and classify digital assets, says Bock & Company principal Geoffrey Bock, who has more than 25 years of experience working with content management issues, is that we are trained to deal with text, rather than pictures, and that presents a problem when we try to classify non-textual assets. "The thing that makes digital asset management different from document management or text asset management is that we are dealing with a visual and spatial component as opposed to a textual component. One of the problems is that when we store things in our computer systems, we are still primarily in a textual space. We are used to dealing with lists of information, not with three-dimensional spatial information. So it's a whole different way of thinking about things," he says.

Bock asserts that if you have a professional staff to help you build the system, it can be a smoother process, but many companies rely on untrained staff to add the metadata, which can lead to inconsistent application. "Metadata is textual, so then the problem is who defines and maintains metadata and how do we assign metadata to assets?" Bock says. "If we approach it from an information-science point of view and we have a librarian and that person knows how to assign this data, and other people know what data is out there, then we have an effective digital asset management system. The problem is that many of us are not trained as specialists on how to catalog assets, nor do we think in the terms of professionals," Bock says.

Susan Worthy, VP of marketing at DAM vendor ClearStory, developer of ActiveMedia 7, agrees and says that applying good metadata is a real challenge. "When you do have a lot of content in the system, as with video, where you don't have the luxury of text, you have a very complex layered file where it's non-textual, so how do you find that? And that's where metadata becomes critical. And not just at the high level of the file itself, but also when you get into the sub-components of the file, because you may need to find that content based on an element within the content, which makes it an even more complex challenge," Worthy says.

Designing a Metadata System
Metadata can come from several different places and exist on different levels, according to Worthy. For example, she says, there is metadata about the status of the file that end users may not typically need to see. There may be metadata that users add as part of the process of ingesting the asset into the system such as a picture caption, date taken, and photographer, and rendition information such as whether this is a high resolution image or the master. There may be metadata that the software extracts from the asset file and applies automatically. In addition, there may be keywords applied to the asset, and these keywords and public metadata information are the basis for locating these files in a search component later on.

Regardless of the source of the metadata, when you build a metadata system, you have to put some serious thought into the structure of the system, and the starting point has to be the end users of the system, says Chris Porter, director of content operations at Getty Images, an online commercial repository with thousands of images and videos. "The most important thing is to look at the approach from a user perspective and work back from there. And that in itself is a pretty encompassing task because, depending on who is going to use the system, they may have different needs from the system and all of those needs have to be accounted for." 

Porter says a structured approach to metadata is essential whether you are building an internal site or external customer-facing site, like Getty Images. As you design the system, you need to measure how much you want to invest in the metadata creation process based on your company's requirements because, she says, it is expensive to create metadata. "A structured and consistent metadata approach is an important element of both an internal tool and an externally facing Web site," she says. "But if your metadata is supporting your core business--for example, it is powering your customer-facing site, which delivers revenue--then it is extremely important to have a structured and consistent metadata approach." 

Bock says that while there are tools to help you along in this process, there is still no substitute for human intelligence. "Before you turn to tools, it takes a wee bit of imagination and hard work," he says. "By that I mean you need to sit down and figure out for yourself what are your major categories and sub-categories. Which ones are important and how do you organize them? What are the ways you want people to organize information? You need to build a thesaurus, the category trees for your organization, and no amount of software is going to help you overcome the need for human intelligence there."  

Page 1 of 3