Understanding information taxonomy helps build better apps

Information architects grapple with taxonomy, but developers often ignore it--to their own detriment. Understanding information taxonomy is the first step in designing better software from the ground up.

Taxonomy represents the foundation upon which information architecture stands, and all well-rounded developers should have at least a basic understanding of taxonomy to ensure that they can create organized, logical applications. But before diving into the topic of taxonomy, let's look briefly at information architecture. That way, we can view taxonomy in its proper place within a broader field of study.

In their book titled Information Ecology: Mastering the Information and Knowledge Environment, Thomas H. Davenport and Laurence Prusak offer this definition:

"Information architecture, in the broadest sense, is simply a set of aids that match information needs with information resources. A well-implemented architectural design structures information in an organization through specific formats, categories, and relationships."

Information architecture consists of six components, as shown in Figure A. Here's a look at each one.

Information usage patterns provide the context for information architecture. They describe how information is utilized and flows within an organization, as typically illustrated through use cases. Information usage patterns may vary by user profile. For example, the information usage pattern for a CIO may be different from the usage pattern for a Web developer in terms of subjects, level of details, content coverage, etc.

Information access points provide various channels through which an information seeker can locate and retrieve the information needed. Examples include breaking news, site maps, search utilities, popular links, and directory listings of categories.

Taxonomy is the core component of information architecture. As illustrated in the diagram, taxonomy interrelates with all of the other components of information architecture. Derived from analysis of usage patterns and information flow, taxonomy guides the visual design of information navigation and administration in compliance with standards and guidelines. There are two aspects of taxonomy: view and structure.

Administration structure provides the visual design for internal users who are responsible for creating and maintaining information, such as taxonomy categories made available through the content management flow.

Information flow describes how information flows within an organization.

Standards and guidelines govern all the components listed above, dictating how new information should be tagged and cataloged.

Figure A
Taxonomy of information architecture

Taxonomy as a two-piece puzzle
The term taxonomy was originally borrowed from the life sciences discipline, where a plant or animal is placed in a single spot describing its hierarchical relationship to other plants and animals. Taxonomy is essentially a type of conceptual framework. It is not a product and has no direct relationship with sales or revenue. When used in the context of the Internet, taxonomy refers to the effective structuring of content within a defined scope to facilitate easy and accurate access.

Depending on the audience, the definition for taxonomy may need some adjustment. For a speech to a CIO and technical professionals: Taxonomy is a conceptual framework for organizing enterprise or companywide content so our employees, partners, and customers can locate what they need easily.

For conversations with friends and families at a dinner party: Taxonomy is about how to organize, label, and arrange information so that you can find the information you need easily, as demonstrated by Yahoo's directory listing.

There are two camps when it comes to what taxonomy really looks like. Camp A refers to taxonomy as a directory listing or site map. Type A campers usually come from a Web development background. Camp B refers to taxonomy as a classification system, thesauri, and controlled vocabulary. Type B campers usually come from a library and information science background. Having camped in both the Web development and library science communities, I would like to demystify the taxonomy puzzle by breaking it into two pieces: taxonomy structure and taxonomy view. Using the classic model-view-controller framework, taxonomy structure is the model and taxonomy view is the view.

Taxonomic structure
Taxonomy structure is similar to classification systems used in the life sciences field and the library community, such as the National Center for Biotechnology Information (NCBI) taxonomy (see Listing A) and the Library of Congress subject classification system (see Listing B).

Taxonomy structure has the following key characteristics:
  • It is hierarchical. Similar to the classification systems illustrated in Listings A and B, taxonomy structure is multilevel, representing hierarchical relationships between concepts within a defined scope and context.
  • It is used to categorize information. Whereas the Library of Congress classification system is used to classify publications held by a library, taxonomy structure for an organization is used to categorize enterprisewide information.
  • It is an integral part of a content management system. Taxonomy structure should be made available within the content management workflow. An authorized user should be given a hierarchical listing of categories from which he or she can assign labels to content items. The assigned category should then be reviewed as part of the assessment and approval process.

Taxonomy view
Taxonomy view is the visual aspect of taxonomy structure. It presents Web content logically by grouping information into topics so a site visitor can navigate and locate information easily. Examples include Yahoo’s subject directory (see Figure B) and Lands’ End’s categorized topics (see Figure C).

Figure B
Yahoo's subject directory: Source: http://www.yahoo.com

Figure C
Land's End categorized topics: Source: http://www.landsend.com

Taxonomy view has the following key characteristics:
  • It may completely mirror taxonomy structure or may be completely different. Theoretically, an entire taxonomy structure can be exposed on the Web site by providing a multilevel treelike visual interface. Through such an interface, a user can drill down to each category and view a listing of content assigned to the selected category, similar to Yahoo’s subject directory.
  • It may not always be visually presented in a hierarchical format. Lands’ End divides its products into six broad topics: Women, Men, Kids, For the Home, Luggage, and Special Collections. Subtopics for the selected broad topic are listed on the left. The next level of subtopics is presented in the center. The screen in Figure C was generated by first selecting For The Home and then, Home Furnishings And Accents. As illustrated, the navigation structure is not provided in the conventional treelike format. Rather, the multilevel navigation is achieved by using tabs as the primary navigation, the left panel as the secondary navigation, and the main content area as the third level of navigation. This is one of the modern design formats typically seen on major Web sites.
  • There can be multiple views of the same taxonomy structure. Other than model-view-controller, we may also think of taxonomy view as the visual presentation of a structure similar to database view. There may be multiple views for the same database structures. The hierarchical associations of product types on an e-commerce site like Lands’ End can be visually presented in many ways. The ultimate design pattern needs to be based on the information usage pattern of the Web site audience.
  • It is an integral part of Web site design and needs to derive from analysis of the information usage patterns of the target audience. An information architect or a taxonomy expert needs to work closely with graphic artists to codevelop the ultimate site design. Whereas an information architect provides the conceptual navigation model for the Web site, graphic artists are responsible for applying visual treatments to the conceptual model and translating it into the final design.

Taxonomy recap
Here's a summary of the key facts about taxonomy:
  • It is a conceptual framework for organizing information in the form of taxonomy view or taxonomy structure.
  • It is tightly related to a defined scope and context. For example, taxonomy for Lands' End would be different from taxonomy for Yahoo because of the different business context, content, and target users.
  • It is a component of information architecture.