Information architecture—CNET Networks

Phil Gochenour

When most people think of a library, they think of books arranged on shelves receding into the distance. What they don’t see is that the books themselves are simply information storage devices that structure data in very specific ways, and that the shelves are merely one method of storing information that makes it accessible and retrievable to a user.

As a technical communicator and information architect, I am involved in creating both the structures for storing information and the systems that make it retrievable. In other words, I help design the books, the library that houses them, and the catalog systems that manage them.

Structuring information means thinking about the same elements associated with any written work: a title that identifies it, a table of contents that tells us what is in it, paragraphs that lead the reader from one idea to the next, and indexes that enable us to locate specific items within it. When we shift from the material world of books to the immaterial world of online content systems (whether we are talking about Wikis, blogs, Web sites, or content that resides in a database), we are still dealing with these same elements.  But we need to consider that they have just as much significance for machines as they do for humans; search engines, for example, will build their indexes based on the document titles and keywords within or associated with the text. We also have to consider that humans process information on a screen differently than they do when it is on a page. Both machines and humans scan, rather than read, online content, and structural elements are the key means by which they identify, classify, and process a specific piece of information.  Most human users, for example, are looking for something specific when they come to a Web page – how to do something, whom to contact, where to get something, etc.  The goal is to structure the information on a page in such a way that the human user can quickly find the thing that they want, or alternative sources for it. This may involve breaking information up into several chunks that are connected by links, shortening paragraphs to bullet points, or thinking about how humans scan Web pages and placing the most important information in the most prominent location. Understanding the best way to structure information is a primary concern for the information architect.

Managing the information retrieval system also involves familiar, yet slightly different, ways of thinking about the library and its contents. For example, all libraries structure the information within them based on taxonomy; the taxonomic system that is most familiar to all of us is the Dewey Decimal System, which provides a way of classifying any piece of information that we want to put into the system.  An immaterial library relies on taxonomies as well, though they need to be developed with the specific information that the system will be handling in mind. For example, one of my recent contract undertakings was to organize documents related to software development projects on a file server, and to create a structure that could be duplicated for other projects.  In my current work with a corporate Wiki, I help different groups develop an information architecture for their individual libraries that allows both internal and external users to locate specific types of documents. In both cases, the challenge is to develop a classification system that works not just for existing documents, but also for ones that might be added in the future.

Finally, there is the design of the system retrieving this information and making it available to users, and once again we can look at a familiar object, the card catalog. Card catalogs were basically very simple databases, with three key fields for each item – author, title, subject – and additional identifying information (what we would call “metadata” in today’s parlance). It was accessed by hand, and the cards served as proxies for the items they cataloged. Today’s content management systems are really no different than the card catalog, though the distinction between the item and its catalog entry has become blurred, and the metadata that we use to catalog items will be crafted to reflect both the information in the item and the system that we are using to retrieve it. If we are designing Web pages that are to be indexed by certain search engines, for example, we may include certain keywords in the head portion of the page, while Google looks at the title of the page, keywords in the first paragraph, and the number of incoming links to that page. The Wiki I help manage is based on the TWiki platform, which allows for the creation of complex search strings that retrieve metadata from forms attached to TWiki topics. Tags are becoming increasingly important as tools for indexing and retrieving online content, and content management systems like Alfresco, Drupal, or Joomla have their own internal systems for cataloging and retrieving information.

Though the term “knowledge management” has fallen into disrepute over the past few years and is sometimes regarded as business jargon, I feel that it is a accurate way to describe the overall objective of library and information science. In my case, I look at a corporation as being made up of hundreds or thousands of individuals, each of whom carries specialized knowledge about their work and the company. We could say that each person represents a small collection of data that, when brought into a consistent structure, becomes collective information. When others have the ability to access that information and incorporate it into their own work, it becomes knowledge.  In more poetic moments I believe that my job is to help the company think about itself and what it does, and that I do this by creating the structures and systems that help turn random data into useful knowledge.